U.S. patent number 5,696,774 [Application Number 08/565,973] was granted by the patent office on 1997-12-09 for digital signal recording device, digital signal playback device, and digital signal decoding device therefor.
This patent grant is currently assigned to Mitsubishi Denki Kabushiki Kaisha. Invention is credited to Sadayuki Inoue, Ken Onishi, Junko Shinohara, Tatsuo Yamasaki.
United States Patent |
5,696,774 |
Inoue , et al. |
December 9, 1997 |
Digital signal recording device, digital signal playback device,
and digital signal decoding device therefor
Abstract
To reduce the undetected errors without degradation of the burst
error correction capability, in error correction decoding of the
digital signal with three-dimensional error correction codes
appended thereto. For correcting errors by applying error
correction codes in three directions of a recording direction,
perpendicular direction, and direction of tracks, whether a burst
error has occurred is checked, and an error-correction decoding
algorithm is switched between one used when a burst error has
occurred, and one used when no burst error has occurred.
Inventors: |
Inoue; Sadayuki (Tokyo,
JP), Shinohara; Junko (Tokyo, JP),
Yamasaki; Tatsuo (Tokyo, JP), Onishi; Ken (Tokyo,
JP) |
Assignee: |
Mitsubishi Denki Kabushiki
Kaisha (Tokyo, JP)
|
Family
ID: |
27301979 |
Appl.
No.: |
08/565,973 |
Filed: |
December 1, 1995 |
Foreign Application Priority Data
|
|
|
|
|
Dec 1, 1994 [JP] |
|
|
6-298461 |
Mar 31, 1995 [JP] |
|
|
7-075893 |
May 10, 1995 [JP] |
|
|
7-111879 |
|
Current U.S.
Class: |
714/755;
375/E7.279; 386/E9.013; 714/761; G9B/20.053; G9B/20.054 |
Current CPC
Class: |
G11B
20/1833 (20130101); G11B 20/1866 (20130101); H04N
9/8042 (20130101); H04N 19/89 (20141101); G11B
5/0086 (20130101); G11B 5/09 (20130101); G11B
15/03 (20130101); G11B 15/1875 (20130101); H04N
5/78266 (20130101); H04N 5/783 (20130101); H04N
9/8227 (20130101) |
Current International
Class: |
G11B
20/18 (20060101); H04N 9/804 (20060101); H04N
7/64 (20060101); G11B 15/18 (20060101); G11B
15/03 (20060101); H04N 9/82 (20060101); H04N
5/783 (20060101); H04N 5/7824 (20060101); H04N
5/7826 (20060101); G11B 020/18 () |
Field of
Search: |
;371/37.4,38.1,40.1 |
References Cited
[Referenced By]
U.S. Patent Documents
Foreign Patent Documents
|
|
|
|
|
|
|
3523249 |
|
Jan 1986 |
|
DE |
|
4224326 |
|
Feb 1994 |
|
DE |
|
Other References
HDTV '93, A Recording Method of ATV data on a Consumer digital VCR,
International Workshop on HDTV '93, Oct. 26-28, 1993, Proceedings,
vol. II ..
|
Primary Examiner: Baker; Stephen M.
Attorney, Agent or Firm: Birch, Stewart, Kolasch &
Birch, LLP
Claims
What is claimed is:
1. A digital signal playback device for playback from a magnetic
tape on which record data is recorded, after arranging in two
dimensions, consisting of a recording direction and perpendicular
direction, the record data to be recorded in predetermined areas on
tracks formed helically on magnetic tape, collecting a plurality of
record data each arranged in the two dimensions to structure a data
block, appending a third error correction code in a third direction
including a direction of depth of the data block, and appending
first and second error correction codes in the recording direction
and perpendicular direction to the record data containing the third
error correction code, said digital signal playback device
comprising:
first error-correction decoding means for performing error
correction on the playback data using the first error correction
code;
burst error detecting means for detecting a burst error that has
occurred in a playback signal during playback;
second error-correction decoding means for performing error
correction on the playback data using the second error correction
code; and
third error-correction decoding means for performing error
correction on the playback data using the third error correction
code;
wherein when a burst error is detected by the burst error detecting
means, an error-correction decoding algorithm followed by at least
one of the second and third error-correction decoding means is made
different from the one followed thereby when no burst error is
detected.
2. The digital signal playback device as set forth in claim 1,
wherein the burst error detecting means recognizes occurrence of a
burst error on the basis of the continuity of error-detection flags
set as a result of application of the first error correction
code.
3. The digital signal playback device as set forth in claim 1,
wherein the burst error detecting means compares the output level
of a playback signal sent from a head with a predetermined level in
normal playback mode, and when the output level of a playback
signal is not higher than the predetermined level continuously for
a predetermined period of time or longer, the burst error detecting
means recognizes occurrence of a burst error.
4. The digital signal playback device as set forth in claim 1,
wherein, when a burst error is detected by the burst error
detecting means, an error-correction decoding algorithm followed by
the second error-correction decoding means is made different from
the one followed thereby when no burst error is detected.
5. The digital signal playback device as set forth in claim 4,
wherein when the burst error detecting means detect a burst error,
error-correction decoding based on the second error correction code
is not executed.
6. The digital signal playback device as set forth in claim 1,
wherein, when a burst error is detected by the burst error
detecting means, an error-correction decoding algorithm followed by
the third error-correction decoding means is made different from
the one followed thereby when no burst error is detected.
7. The digital signal playback device as set forth in claim 6,
wherein when the burst error detecting means detects a burst error,
the third error-correction decoding means uses error-detection
flags set as a result of application of the first error correction
code in performing error correction on data residing on a track
with regard to which the burst error is detected.
8. The digital signal playback device as set forth in claim 6,
wherein, a maximum correctable number of erasures in correcting
errors by means of the third error-correction decoding means is
different depending on whether or not the burst error detecting
means detects a burst error.
9. An error-correction decoding device for performing
error-correction using first to third error correction codes on
received or reproduced digital data, having been transmitted or
recorded, after arranging transmission or recording digital data in
two dimensions, consisting of a transmitting or recording
direction, and perpendicular direction, collecting a plurality of
digital data each arranged in the two dimensions to structure a
data block, applying the third error correction code in a third
direction including a direction of depth of the data block,
applying the first and second error correction codes in the
transmitting or recording direction, and perpendicular direction to
the digital data containing the third error correction code, and
then transmitting or recording the digital data, said device
comprising:
burst error detecting means for detecting a burst error that has
occurred in the received or reproduced digital data;
first error-correcting means for performing error correction on the
received or reproduced digital data using the first error
correction code;
second error-correcting means for of performing error correction on
the received or reproduced digital data using the second error
correction code;
third error-correcting means for performing error correction on the
received or reproduced digital data using the third error
correction code; and
control means for controlling such that, when a bust error is
detected, at least one of the second and third error correction
means uses an error correction decoding algorithm different from
the one used when no burst error is detected.
10. The error correction decoding device as set forth in claim 9,
wherein said burst error correction means recognizes occurrence of
a burst error on the basis of the continuity of error-detection
flags set as a result of the error correction by the first
error-correcting means.
11. The device according to claim 9, wherein said control means
controls such that, when a burst error is detected, said second
error-correcting means uses an error correction decoding algorithm
different from the one used when no burst error is detected.
12. The error correction decoding device as set forth in claim 11,
wherein, when said burst error detecting means detects a burst
error, the second error-correcting means does not execute the error
correction.
13. The error correction decoding device according to claim 9,
wherein said control means controls such that, when a burst error
is detected said third error-correcting means uses an error
correction decoding algorithm different from the one used when no
burst error is detected.
14. The error correction decoding device as set forth in claim 13,
wherein, when said burst error detecting means detects a burst
error, error-detection flags set as a result of the error
correction by the first error-correcting means are used to perform
error correction on the digital data residing on a plane on which
the burst error is detected.
15. The error correction decoding device as set forth in claim 13,
wherein, said third error-correcting means uses different maximum
correctable number of erasures depending on whether the burst error
detecting means does or does not detect a burst error error.
16. A digital signal recording device in which intra-frame or
-field coded or inter-frame or -field coded digital video and audio
signals input in the form of transport packets are
transparent-recorded, comprising:
data separating means for separating the intra-frame or -field
coded digital video signal from the transport packets;
special playback data generating means for reconfiguring the
intra-frame or -field coded digital video signal having been
separated at the data separating means, to generate special
playback data; and
recording data generating means for re-configuring the data of the
transport packets having been input, to generate sync block
formats;
wherein
the area for recording the special playback data and error
correction code is provided on a scanning trace of a head scanning
the recording media at a predefined fast playback speed,
one of the error correction code appended to the special playback
data or the error correction code appended at least the data output
from the recording data generating means is recorded in the area
for recording the error correction code, and
an identification signal identifying the error correction code
recorded in the error correction code recording area is appended to
the recording signal, and recorded.
17. The digital signal recording device according to claim 16,
wherein the special playback data generated by the special playback
data generating means is generated in the form of the transport
packets having been input, and, at the time of recording, converted
into sync block format like the transport packets having been
input, and then recorded.
18. The digital signal recording device according to claim 16,
wherein the recording area for the error correction code is
disposed adjacent to the special playback data recording area.
19. The digital signal recording device according to claim 16,
wherein when the error correction code is appended to the output of
the recording data generating means, the output of the recording
generating means and the output of the part or entirety of the
special playback data generating means, to configure the data
block.
20. The digital signal recording device according to claim 16
wherein when the sync block formats are generated by the recording
format generating means, m transport packets are used to generate n
lines of sync blocks of data, and the recording data is so disposed
that the above-mentioned n lines of sync blocks of data output from
the recording format generating means do not extend beyond one
recording track.
21. The digital signal recording device according to claim 20,
wherein the identification signal for identifying the data recorded
in the error correction code recording area is transmitted using a
recording area provided at the data recording part which remain
when the transport packets are converted into the sync block
formats at the recording data generating means.
Description
BACKGROUND OF THE INVENTION
The present invention relates to a digital video tape recorder
(hereinafter referred to as a digital VTR) having a track format in
which digital video and audio signals are recorded in predefined
areas on oblique tracks, wherein digital video and audio signals
are input in the form of a bit stream, and the bit stream is
recorded.
The present invention also relates to a digital signal playback
device such as a digital VTR or digital disk player for reproducing
the digital data having been transmitted or recorded after
arranging transmitted or recorded digital data in two dimensions,
that is, in a transmitting or recording direction and perpendicular
direction, collecting a plurality of digital data each arranged in
the two dimensions to structure a data block, applying a third
error correction code in a third direction including a direction of
depth of the data block; and applying first and second error
correction codes in the transmitting or recording direction and
perpendicular direction to the digital data containing the third
error correcting code, and in particular to an error-correction
decoding for a digital data receiver for receiving data that has
been encoded according to the above procedure.
FIG. 1 shows a pattern of tracks for a conventional typical
consumer digital VTR. In the drawing, helical tracks are formed on
magnetic tape. One track is divided into two areas: a video area in
which a digital video signal is recorded; and an audio area in
which a digital audio signal is recorded.
There are two methods for recording video and audio signals in this
type of consumer digital VTR. One of the methods is what is called
a baseband recording method, wherein analog video and audio signals
are received, subjected to high-efficiency coding, and recorded at
a lowered data rate. The other one is what is called a transparent
recording method for recording a bit stream transmitted in digital
form.
For recording an advanced television (hereinafter ATV) signal that
has been discussed in the U.S.A., or a digital video broadcasting
(hereinafter DVB) signal that has been studied in Europe, the
latter transparent recording method is suitable. This is because
the ATV signal or DVB signal is a signal that has already been
compressed in digital form and therefore need not be treated by a
high-efficiency encoder and decoder. Moreover, since the ATV or DVB
signal can be recorded as it is, picture quality will therefore not
deteriorate. Drawbacks are that the ATV or DVB signal is quite
sensitive to an error occurring during playback and the picture
quality during special playback, such as fast, still or slow
playback is poor. In particular, when an error occurs during fast
playback, the error propagates over several frames. This results in
deterioration in quality of reproduced pictures. If a rotary head
records a bit stream on oblique tracks as it is, pictures are
hardly reproduced during special playback.
According to a basic specification for a prototype of a consumer
digital VTR, one frame of an image is recorded in video areas on
ten tracks, in the standard definition (SD) mode, when the
recording rate of a digital video signal is 25 Mbps and the field
frequency is 60 Hz. If the data rate of the ATV signal is set to 17
to 18 Mbps, the ATV signal can be recorded in the SD mode using the
transparent method.
A recording format according to the SD specification (hereinafter
the SD standard) that is the basic specification for the prototype
of a consumer digital VTR will be described below. The SD standard
stipulates that when the field frequency is 60 Hz, a video signal
representing one frame is recorded on ten tracks, as mentioned
above. When the field frequency is 50 Hz (under the PAL or SECAM
system), a video signal representing one frame is recorded on
twelve tracks. FIG. 2 shows a recording format for one track under
the SD standard. As illustrated, the SD standard stipulates that a
video area is composed of a total of 149 sync blocks: 135 sync
blocks constituting an area in which image data is recorded; three
sync blocks constituting an area in which VAUX data is recorded,
and eleven sync blocks constituting an area in which error
correction codes are recorded perpendicularly. Numerical characters
written on the left of the video area denote sync block addresses
within one track. The VAUX data is recorded at sync block addresses
19, 20, and 156. The image data is recorded at sync block addresses
21 to 155.
FIG. 3A and FIG. 3B show the arrangement of error correction codes
appended to image data and audio data under the foregoing SD
standard. According to the SD standard, a (85, 77, 9) Reed-Solomon
code (hereinafter a C1 check code) is used as an error correction
code applied to image data in the recording direction, and a (149,
138, 12) Reed-Solomon code (hereinafter a C2 check code) is used as
an error correction code applied in the perpendicular direction.
Similarly, the (85, 77, 9) Reed-Solomon code (C1 check code),
identical to that for the image data, is used as an error
correction code applied to audio data in the recording direction,
and a (14, 9, 6) Reed-Solomon code (hereinafter a C3 check code) is
used as an error correction code applied to the audio data in the
perpendicular direction. FIG. 4 shows the structure of one sync
block. As shown in FIG. 4, one sync block is 90 bytes long. Five
bytes at the head are used to record a sync pattern and an ID
signal. Eight bytes at the tail are used to record error correction
codes (C1 check code).
A type of digital VTR for transparent recording the above-mentioned
ATV signal on magnetic tape using the track format stipulated by
the SD specification is disclosed in "A Recording Method of ATV
data on a Consumer Digital VCR" presented at the "International
Workshop on HDTV '93" held in Ottawa, Canada on Oct. 26 through 28,
1993. The contents of the method will next be described as a prior
art example. As mentioned above, the data rate of the ATV signal
ranges from 17 to 18 Mbps. A digital VTR conforming to the SD
standard can record the ATV signal in transparent mode.
A bit stream constituting an ATV or DVB signal is conforming to the
MPEG2 system. That is, image data is encoded by intra-frame or
-field coding in which data is encoded using correlation within one
frame, or by inter-frame or -field coding (motion compensated
prediction) in which data compression is performed using
correlation of data between different frames (or fields). For
reproducing data compressed according to the MPEG2 system, when
errors occur in a playback signal, the errors propagate over
several frames until the next intra-frame coded data is reproduced.
This is because only the intra-frame coded data can be decoded
independently without the need of referencing any other frame.
A digital VTR disclosed on the above technical report will be
described below. The disclosed technology is concerned with a
recording format that is used for recording the ATV signal on
magnetic tape and useful in realizing fast playback which is one of
the aforesaid problems. The recording format will be described
briefly. FIG. 5A and FIG. 5B show the scanning traces of rotary
heads in normal playback and fast playback modes in a conventional
digital VTR. In the drawing, adjoining tracks are recorded
alternately by rotary heads having different azimuths. In normal
playback mode, the tape travel speed (which may hereinafter be
referred to simply as a tape speed) is identical to that for
recording. The rotary heads can trace the recording tracks as shown
in FIG. 5A. The tape speed in fast playback mode is different from
that for recording. The rotary heads traverse several tracks, and
reproduce fractions of data from the parts of the tracks of the
identical azimuth. FIG. 5B shows a scanning trace at a 5.times.
feed. For example, when a bit stream conforming to the MPEG2 system
is recorded sequentially on tracks, only the data that is
intra-frame coded and separated from a signal reproduced
intermittently is reproduced in fast playback mode. The intra-frame
coded data alone is used to reconstruct a picture. The reproduced
part on the screen are not contiguous to one another, and
reproduced fractions of data are spread over each screen. Moreover,
since the bit stream is variable-length coded, it is not sure that
all the data (pixels) on the screen are updated cyclically, and
certain portions of the screens may not be updated for a prolonged
period of time. As a result, picture quality attained in fast
playback mode is not satisfactory and unacceptable for consumer
digital VTRs.
FIG. 6 is a block diagram showing the configuration of a
conventional digital VTR permitting fast playback. Herein, the
video area of each track is divided into a main area in which a bit
stream of an ATV signal is recorded and a duplication area in which
an important portion (high-priority (hereinafter HP) data) of a bit
stream used for reconstructing a picture in fast playback mode is
recorded. In fast playback mode, only the sync blocks containing
intra-frame coded data are valid. The intra-coded data is therefore
recorded in the duplication areas. For further reducing the amount
of data, low-frequency components are extracted from the
intra-frame coded data and recorded as HP data. In FIG. 6,
reference numeral 1 denotes a bit stream input terminal. 2 denotes
a bit stream output terminal. 3 denotes an HP data output terminal.
4 denotes a variable-length decoder. 5 denotes a counter. 6 denotes
a data extracting circuit. 7 denotes an end-of-block (hereinafter
EOB) appending circuit.
An MPEG2-conforming bit stream is fed through the input terminal 1,
supplied as it is through the output terminal 2, and recorded
sequentially in the main areas. The bit stream from the input
terminal 1 is also fed to the variable-length decoder 4. The syntax
of the MPEG2-conforming bit stream is analyzed and an intra picture
image is detected. A timing signal is produced by the counter 5.
Low-frequency components are extracted from all data blocks of an
intra image by the data extracting circuit 6. An EOB is appended by
the EOB appending circuit 7. Thus, HP data is constructed and
recorded in the duplication areas.
FIG. 7 schematically illustrate normal playback and fast playback
in the conventional digital VTR. In normal playback, all bit
streams recorded in the main areas are reproduced and fed to an
MPEG2-conforming decoder installed externally to the digital VTR.
The HP data is discarded. By contrast, in fast playback mode, only
the HP data recorded in the duplication areas is acquired and fed
to the decoder. The bit streams in the main areas are
discarded.
Next, the locations of main areas and duplication areas in each
track will be described. FIG. 8 shows a scanning trace of a rotary
head in fast playback mode. Under the conditions that the tape
speed is set to an integral multiple of a normal tape speed and
phase lock is attained, the scanning head is synchronized with
tracks of identical azimuth. The locations of data to be reproduced
are fixed. In FIG. 8, assuming that part of the playback signal
having output levels that exceed -6 dB will become valid playback
data, the part of the playback signal which will be valid playback
data during one scan of the head are depicted as crosshatched areas
in FIG. 8. FIG. 8 shows 9.times. speed playback. That is to say, it
is ensured that the part of a playback signal depicted as
crosshatched areas in FIG. 8 is read out (i.e., obtained as valid
data) at the 9.times. speed. The HP data should therefore be
recorded in these areas. However, at other multiple speeds, reading
of data from these areas is not ensured. It is therefore necessary
to select the areas so that reading is ensured at several tape
speeds.
FIG. 9 is an explanatory diagram concerning areas (overlap areas)
on tape that are scanned in common at various conventional fast
playback speeds, e.g., at three tape speeds. Several overlap areas
are scanned at each of the tape speeds. Duplication areas are
selected from among the overlap areas, so as to ensure reading of
the HP data recorded in the duplication areas at the different tape
speeds. FIG. 9 is concerned with fast tape feed at 4.times.,
9.times., and 17.times. speeds. The scanned areas in FIG. 9 are
also scanned during fast tape feed at -2.times., -7.times., and
-15.times. speeds.
Not all the overlap areas can however be traced by the head at
different tape speeds. This is because the number of tracks
traversed by the head differs from one tape speed to another. It is
also necessary that the head can start scanning at any identical
azimuth track. FIG. 10 shows the scanning traces of a rotary head
at 5.times. and 9.times. speeds in the conventional digital VTR. In
the drawing, areas 1, 2, and 3 are selected from among the overlap
areas at the 5.times. and 9.times. speeds. When the same HP data is
recorded repetitively on nine tracks, the HP data can be read out
at either of the 5.times. and 9.times. speeds.
FIG. 11 shows the scanning traces of two rotary heads in 5.times.
speed playback mode in the conventional digital VTR. As apparent
from the drawing, when HP data is recorded repetitively on the
number of tracks identical to a speed multiplier (ratio of the fast
playback tape speed to the normal playback tape speed), the HP data
can be read out by the heads that are synchronized with the
identical azimuth tracks. In other words, when the HP data is
duplicated repetitively on the number of tracks identical to the
speed multiplier of the maximum multiple tape speed for the fast
playback, the HP data can reliably be read out in either of forward
and reverse directions at various tape speeds.
FIG. 12 shows the structure of a track for the conventional digital
VTR. In the illustrated example, a track has main areas and
duplication areas. As far as a consumer digital VTR is concerned, a
video area of each track is composed of 135 sync blocks. The main
areas are composed of 97 sync blocks. The duplication areas are
composed of 32 sync blocks. The duplication areas include overlap
areas for the 4.times., 9.times., and 17.times. speeds in FIG. 9.
In this case, a data rate for the main areas is about 17.46 Mbps.
As for the duplication areas, since the same data is recorded
repetitively 17 times, the data rate is about 338.8 kbps.
The conventional consumer digital VTR has the aforesaid
configuration, and special playback data is repetitively recorded
in the duplication areas. The recording rate of the special
playback data is therefore low. In slow playback or fast playback
mode, especially, the quality of reproduced pictures is
unsatisfactory. For example, when intra-frame coding is performed
twice per second, the data rate of intra-coded data alone in an ATV
signal is estimated to be about 3 Mbps, but the conventional VTR
can merely achieve recording at about 340 kbps. The quality of
reproduced pictures is therefore quite poor.
Moreover, as mentioned above, for an ATV or DVB signal, data
compression is performed using a technique based on
motion-compensated prediction. The compressed data is composed of
intra-data (intra-frame or -field coded data) whose playback data
alone makes it possible to restore a picture, and inter-data
(inter-frame or -field coded data) which can be used in combination
with reference-field or -frame data to restore a picture. When
errors occur in the playback data, the errors propagate over a
plurality of fields or frames in the case of an ATV signal. This
results in an unpleasant view. When the digital VTR conforming to
the SD standard is used as a storage media for storing data or
programs for a computer or the like, it is desired to apply a more
powerful error correction code in order to restore (achieve error
correction of) data that has not been reproduced because of
drop-outs caused by flaws or dust on magnetic tape.
Furthermore, because the rotary heads cross the recording tracks
obliquely during special playback (fast playback, slow playback,
still playback or the like), the playback signal is produced
intermittently from each track. It is therefore not possible to
structure the error correction block (video data) shown in FIG. 3A.
Accordingly, the playback data is subjected to error correction
with the C1 check code only, during special playback.
When error correction by means of C1 check code is used, and the
symbol error rate is 0.01, the error correction probability will be
1.56.times.10.sup.-3 and one error per 8 sync blocks will be
detected. In particular, playback output is not stable during
special playback, and the symbol error rate often exceeds 0.01. The
recorded data is variable-length encoded, so that when error occurs
the subsequent data cannot be used, and the playback picture
quality deteriorate. The undetected error rate will also be as high
as 7.00.times.10.sup.-8.
SUMMARY OF THE INVENTION
The invention has been achieved to solve the problems discussed
above.
An object of the present invention is to improve the error
correction capability in normal playback and to improve the quality
of reproduced pictures.
Another object of the invention is to improve error correction
capability in fast playback as well as in normal playback, and to
improve playback picture quality during fast playback through
increase in the recording rate of the fast playback data.
According to one aspect of the invention, there is provided a
digital signal playback device for playback from a magnetic tape on
which record data is recorded, after arranging in two dimensions
consisting of a recording direction and perpendicular direction,
the record data to be recorded in predetermined areas on tracks
formed helically on magnetic tape, collecting a plurality of record
data each arranged in the two dimensions to structure a data
appending a third error correction code in a third direction
including a direction of depth of the data block, and appending
first and second error correction codes in the recording direction
and perpendicular direction to the record data containing the third
error correction code.
First error correction means corrects or detects errors occurring
during playback, in normal playback mode.
Burst error detecting means detects a burst error occurring in a
playback signal. As for playback data whose errors have been
corrected or detected by the first error-correction means, second
error-correction decoding means uses the second error correction
code to correct errors detected by the first error-correction
decoding means. The second error-correction decoding means also
corrects errors undetected by the first error correcting means.
In a first embodiment, when the burst error detecting means detects
a burst error, the error-correction decoding algorithm followed at
least by the second error-correction decoding means to correct or
detect errors in playback data is made different from the one
followed thereby when no burst error is detected. As for playback
data whose errors have been corrected or detected by the second
error-correction decoding means, the third error-correction
decoding means uses the third error correction code to correct the
errors detected by the second error-correction decoding means. (The
third error correction detecting means also corrects errors
undetected by the second error-correction decoding means.) Because
the error-correction decoding algorithm followed by the second
error-correction decoding means when a burst error occurs is
changed from the one followed thereby when no burst error occurs,
occurrence of errors undetected by the second error-correction
decoding means (miscorrection, i.e., decoding error) in the event
of a burst error can be reduced. Miscorrection by the third
error-correction decoding means can therefore be reduced.
Especially in a digital signal playback device for reproducing data
that has been subjected to inter-frame (or -field) coding such as
motion-compensated prediction typified by the MPEG2 system, image
quality deterioration due to undetected errors can be prevented.
Moreover, errors can be detected reliably, and interpolation can
therefore be achieved satisfactorily. When the digital signal
playback device is used as an external storage device for a
computer or the like, errors can be detected reliably by the third
error-correction decoding means. An appropriate remedy such as
re-transmission of data can therefore be effected as may be
required.
In a second embodiment, the playback data having been subjected to
error correction and detection by the second error-correction
decoding means is then supplied to the third error-correction
decoding means, which uses the third error correction code to
correct the errors detected by the first or second error correcting
means. (The third error-correction decoding means also corrects
errors undetected by the first or second error correcting means.)
When the burst error detecting means detects a burst error, an
error-correction decoding algorithm followed at least by the third
error-correction decoding means to correct or detect errors in the
playback data is made different from the one followed thereby when
no burst error is detected. The error-correction decoding algorithm
followed by the third error correcting means when a burst error
occurs is changed from the one followed thereby when no burst error
occurs. Accordingly, in the event of a burst error, the error
correction capability of the third error correction code can be
exhibited fully for error correction. When no burst error occurs,
occurrence of errors undetected by the third error correction codes
(miscorrection) can be prevented substantially perfectly.
The object of applying the third error correction code to data on a
storage media in a digital VTR or the like is to correct a long
burst error stemming from drop-outs. When a burst error occurs, the
error correction capability of the third error correction code can
be utilized fully to correct errors in reproduced digital data. As
for data in which no burst error occurs, occurrence of errors
undetected by the third error correction code can be prevented
satisfactorily by upgrading error detection capability. In
particular, with regard to the code configuration which will be
described in connection with Embodiments 1 to 3, it can happen that
the result of the third error correction code (C4 check code) has a
greater undetected error rate than the result of the second error
correction code (C2 check code) at parts where the symbol error
rate is poor (where the minimum hamming distance of the C4 check
code is shorter than that of the C2 check code). but in such a
case. when control is performed as mentioned above, the errors in
reproduced digital data can be corrected reliably without an
increase in the number of errors undetected by the third error
correction code.
According to either of the first and second embodiments described
above, in a digital signal playback device for reproducing data
that has been subjected to inter-frame (or -field) coding such as
motion compensated prediction typified by the MPEG2 system, image
quality deterioration due to undetected can be prevented. Moreover,
errors can be detected reliably. and interpolation can therefore be
achieved satisfactorily. The third error-correction detecting means
can detect errors reliably even when the digital signal playback
device is used as an external storage device for a computer or the
like. An appropriate remedy such as re-transmission of data can
therefore be effected as may be required.
In either of the first and second embodiments described above, it
may be so arranged that when detecting a burst error, the burst
error detecting means recognizes occurrence of a burst error on the
basis of the continuity of error-detection flags set as a result of
application of the first error correction code. A burst error can
therefore be detected reliably. In particular, influence by the
characteristics of a magnetic recording or the like is eliminated.
Individual digital signal playback devices such as digital VTRs
need not be adjusted one by one. The detection of the burst error
can be achieved by performing numerical calculation on playback
data. The adjusting process in assembly can therefore be
simplified.
In either of the first and second embodiments described above, it
may be so arranged that when detecting a burst error in normal
playback mode, the burst error detecting means compares the output
level of a playback signal sent from a head with a predetermined
level. When the output level of the playback signal is not higher
than the predetermined level continuously for a predetermined
period of time or longer, occurrence of a burst error is
recognized. A burst error can therefore be detected from the
playback signal reliably. Moreover, if error-correction decoding
algorithm (especially, the first and second error correction codes)
is switched from one to another, or like action is taken, according
to the output level of the playback signal, error-correction
decoding can be achieved more effectively.
In the first embodiment described above, it may be so arranged that
if the burst error detecting means detects a burst error, the
second error-correction decoding means does not perform
error-correction decoding using the second error correction code.
It is possible to prevent miscorrection by the second error
correction code. (In particular, when a burst error occurs, errors
detected by the first error correction code are almost all false.
When such data which are mostly erroneous is subjected to error
correction, error correction using the second error correction code
is hardly successful, and rather the number of miscorrection
increases, and error correction using the third error correction
code cannot be expected.
In the second embodiment described above, it may be so arranged
that the burst error detecting means detects a burst error, when
the third error-correction decoding means performs error correction
using the third error correction code, errors in data residing on a
plane on which the burst error is detected are corrected using
error-detection flags set as a result of application of the first
error correction code. Accordingly, in the event of a burst error,
the burst error can be corrected reliably by the third error
correction code.
If error correction is performed using the results of error
correction based on the second error correction code in the event
of a burst error, almost all the data in the track where a bust
error has occurred will be error detection, in the case of the code
configuration which will be described in connection with Embodiment
1 or 2. It is therefore not possible to correct the burst errors by
means of the error correction using the second error detection
flags. By the use of the configuration set forth above, when a bust
error occurs, the burst error can be corrected by fully utilizing
the error correction capability of the third error correction code,
while when random errors occur, the error correction of the third
error correction code can be fully utilized against the random
errors as well, by using the error detection flags set as a result
of the application of the second error correction code. Thus, error
correction can be achieved successfully against either of burst
errors and random errors.
In the second embodiment described above, it may be so arranged
that when the third error correction is executed. an
error-correction decoding algorithm is controlled in such a way
that a maximum correctable number of erasures is varied between
when the burst error detecting means detects a burst error and when
no burst error is detected. Accordingly, when the minimum hamming
distance of the third error correction code is shorter than that of
the second error correction code, if random errors occur, part of
the error correction capability of the third error correction code
can be assigned to error detection. Occurrence of errors undetected
by the third error correction code can therefore be reduced
effectively. For correcting a burst error caused by drop-outs, the
error correction capability of the third error correction code is
used fully. The burst error correction capability of the third
error correction code is therefore utilized fully to correct a
burst error.
According to another aspect of the invention, there is provided an
error-correction decoding device for performing error correction
decoding using first to third error correction codes on received or
reproduced digital data having been transmitted or recorded, after
arranging transmission or recording digital data in two dimensions,
that is, in a transmitting or recording direction and a
perpendicular direction, collecting a plurality of digital data
each arranged in two dimensions to structure a data block, applying
the third error correction code in a third direction including a
direction of depth of the data block, and applying the first and
second error correction codes in the transmitting or recording
direction and perpendicular direction to the digital data
containing the third error correction codes. The error-correction
decoding device comprises burst error detecting means for detecting
a burst error occurring in the received or reproduced digital data,
and first error correcting means for correcting errors in the
received or reproduced digital data using the first error
correction code. Second error correcting means corrects errors in
the received or reproduced digital data using the second error
correction code, for correcting errors in the digital data.
The received or reproduced digital data which have been subjected
to error correction using the second error correction code is then
subjected to error correction using the third error correction
code.
In a first embodiment of this decoding device, when a burst error
is detected at the burst error detecting means, a decoding
algorithm followed at the second error correcting means in order to
correct errors in the digital data is made different from the one
followed when no burst error is detected.
Thus, when a burst error occurs, the error-correction decoding
algorithm followed by the second error-correction decoding means is
changed from the one followed thereby when no burst error occurs
into another. It is therefore possible to reduce occurrence of
errors undetected by the second error correction code
(miscorrection) in the event of a burst error. Miscorrection by the
third error correcting means can therefore be reduced.
When the foregoing error correction codes are used to perform error
correction on moving picture data which undergoes inter-field (or
-frame) coding such as motion-compensated prediction typified by
the MPEG2 system, or programs or data for a computer or the like,
if the foregoing error-correction decoding method is used for
error-correction decoding of picture data, deterioration in picture
quality due to Moreover, errors can be reduced. Moreover, errors
can be detected reliably. Interpolation can therefore be achieved
successfully. When the error-correction decoding is implemented in
an external storage device for a computer or the like, the third
error-correction detecting means can detect errors reliably. An
appropriate remedy such as re-transmission of data can therefore be
effected as may be required. As for a digital signal receiver or
the like, a burst error can be detected irrespective of the
properties of a transmission path or the like. Error-correction
decoding can therefore be accomplished successfully.
In a second embodiment of the decoding device, when the error
correction is applied to the digital data at the third error
correcting means, if a burst error is detected at the burst error
detecting means, a decoding algorithm followed at the third error
correcting means in order to correct the errors in the digital data
is made different from the one followed when no burst error is
detected. Because the error-correction decoding algorithm followed
by the third error correcting means when a burst error occurs is
changed from the one followed thereby when no burst error occurs
into another, the error correction capability of the third error
correction code is exhibited fully for error correction when a
burst error occurs. When no burst error occurs, occurrence of
errors undetected by the third error correction code
(miscorrection) can be eliminated almost perfectly.
The object of applying the third error correction code is to
correct a long burst error caused by drop-outs occurring in a
digital signal playback device (a digital VTR, digital disk player,
CD-ROM or the like). When a burst error occurs, the error
correction capability of the third error correction code can be
utilized fully to correct errors in reproduced digital data. As for
data in which no burst error occurs, occurrence of errors
undetected by the third error correction code can be prevented
satisfactorily by upgrading error detection capability. In
particular, with regard to the code configuration which will be
described in connection with Embodiments 1 to 3, it can happen that
the result of the third error correction code (C4 check code) has a
greater undetected error rate than the result of the second error
correction code (C2 check code) at parts where the symbol error
rate is poor (where the minimum hamming distance of the C4 check
code is shorter than that of the C2 check code), but in such a
ease, when control is performed as mentioned above, the errors in
reproduced digital data can be corrected reliably without an
increase in the number of errors undetected by the third error
correction code.
When the foregoing error correction codes are used to correct
errors in moving picture data which undergoes inter-field (or
-frame) coding such as motion-compensated prediction typified by
the MPEG2 standard, or programs or data for a computer or the like,
if the foregoing error-correction decoding method is used for
error-correction decoding of picture data, deterioration in picture
quality due to undetected errors can be suppressed. Moreover,
errors can be detected reliably. Interpolation can therefore be
accomplished satisfactorily. When the foregoing error-correction
decoding is implemented in an external storage device for a
computer or the like, the third error-correction decoding means can
detect errors reliably. An appropriate remedy such as
re-transmission of data can therefore be effected as may be
required. The error correction capability of the third error
correction code against burst errors can be maintained while the
miscorrection in the event of random errors can be reduced. In a
digital signal receiver or the like, a burst error can be detected
irrespective of the properties of a transmission path or the like.
Error-correction decoding can be accomplished successfully.
In either of the first and second embodiments described above, it
may be so arranged that when a burst error occurring in received or
reproduced digital data is detected at the burst error detecting
means, the number of consecutive error-detection flags set by the
first error correcting means is counted. If the number of
consecutive error-detection flags that are set is not less than a
predetermined value, occurrence of a burst error is recognized. A
burst error can therefore be detected reliably. When the
error-correction decoding method is employed in a digital signal
playback device or the like, influence of the characteristics of a
magnetic: recording or the like is eliminated, so that it is not
necessary to adjust the individual digital signal playback devices
one by one. The detection of the burst error can be achieved by
performing numerical calculation on playback data. The adjusting
process in assembly can therefore be simplified. In a digital
signal receiver or the like, a burst error can be detected
irrespective of the properties of a transmission path or the like.
Error-correction decoding can therefore be achieved
successfully.
In the first embodiment described above, it may be so arranged that
if a burst error is detected at the burst error detecting means,
error-correction decoding using the second error correction code is
not executed at the second error correcting means. It is therefore
possible to prevent miscorrection by the second error correction
code. (In particular, when a burst error occurs, errors detected by
the first error correction code are almost all false. When such
data which are mostly erroneous is subjected to error correction,
error correction using the second error correction code is hardly
successful, and rather the number of miscorrections increases, and
error correction using the third error correction code cannot be
expected. Because of the above-described control, when a bust error
occurs, the burst error correction capability of the third error
correction code is fully utilized to correct the burst error.
In the second embodiment described above, it may be so arranged
that if a burst error is detected at the burst error detecting
means, when errors in received or reproduced digital data are
corrected at the third error correcting means, error-detection
flags set by the first error correction means are used to correct
errors in data residing on a plane on which the burst error is
detected, and error-detection flags set by the second error
correcting means are used to correct errors in digital data
residing on a plane on which no burst error is detected.
Accordingly, in the event of a burst error, the burst error can be
corrected reliably by the third error correction code.
As described above, if error correction is performed using the
results of error correction based on the second error correction
code in the event of a burst error, almost all the data in the
track where a bust error has occurred will be error detection, in
the case of the code configuration which will be described in
connection with Embodiment 1 or 2. It is therefore not possible to
correct the burst errors by means of the error correction using the
second error detection flags. By the use of the configuration set
forth above, when a bust error occurs, the burst error can be
corrected by fully utilizing the error correction capability of the
third error correction code (the error correction decoding is
achieved without degrading the burst error correction capability).
while when random errors occur, the error correction of the third
error correction code can be fully utilized against the random
errors as well, by using the error detection flags set as a result
of the application of the second error correction code. Thus, error
correction can be achieved successfully against either of burst
errors and random errors.
In the second embodiment described above, it may be so arranged
that when a burst error is detected error correction based on the
third error correction code is executed at the third error
correcting means, a maximum correctable number of erasures is
varied between when a burst error is detected in the data block at
the burst error detecting means and when no burst error is
detected. Accordingly, when the minimum hamming distance of the
third error correction code is shorter than that of the second
error correction code, if random errors occur, part of the error
correction capability of the third error correction code can be
assigned to error detection. Occurrence of errors undetected by the
third error correction code can therefore be reduced effectively.
For correcting a burst error caused by drop-outs, the error
correction capability of the third error correction code is used
fully. The burst error correction capability of the third error
correction code is therefore utilized fully to correct a burst
error.
According to a further aspect of the invention, there is provided a
digital signal recording device wherein intra-frame or -field coded
or inter-frame or -field coded digital video and audio signals
input in the form of transport packets are transparent-recorded.
first the intra-frame or field coded digital video signal
(hereinafter referred to as intra-picture signal) is separated from
the transport packets. High-frequency components are removed from
the intra-picture signal having been separated, and the recording
rate is then compressed. The intra-picture signal whose recording
rate has been compressed is re-configured to generate special
playback data. The data of the transport packets having been input
is re-configured, and converted into sync block formats. When the
special playback data is recorded, the special playback data is
disposed on traces on the recording media scanned by a head at a
predefined, fast playback speed. The areas for recording the
error-correction codes are also disposed on the scanning traces of
the head. Either the error-correction code appended to the special
playback data or the error-correction code appended the recording
data obtained by conversion into the sync block format is recorded
in the area for recording the error correction code. The recording
format is configured such that an identification signal identifying
the error correction code recorded in the error correction code
recording area is appended to the recording signal, and recorded.
Accordingly, the limited appended data recording area (special
playback data recording area, error correction code recording area
and the like) can be effectively utilized.
For instance, where the digital signal recording device is used as
a consumer device, and image data such as the above-mentioned ATV
signal (or signal employing a method of compression based on motion
compensated prediction represented by MPEG1 or MPEG2 (method of
compression in which image data is restored by referring to other
image data when decoding one picture of image data)) is recorded,
the error correction code for special playback data, for example,
is recorded in the recording area for the error correction code.
Where the digital signal recording device is one for professional
use, or used as an external recording device for a computer or the
error correction code appended to normal playback data is appended
in the error correction code recording area. Since powerful error
correction code can be appended to the desired data with the
identical recording format, and the identification signal is
recorded in the recording signal, it is possible to perform
compatible playback (playback by the use of another compatible
device) with a digital signal recording/playback device having only
one of the decoders of the above-mentioned error correction codes,
by identifying the type of the error correction code recorded in
the error correction code recording area.
Furthermore, the error correction code recording area is used in
common with the area for recording the error correction code
appended to the normal playback data and the error correction code
appended to the special playback data, and switched depending on
the intended use, the error correction code recording area within
the limited appended data recording area can be reduced, and for
that amount the area for recording the special playback data can be
increased, the recording rate of the special playback data can be
increased, and the picture quality during the special playback can
be improved.
It may be so arranged that where, for instance, the data
transmitted in the transport packets represented by MPEG2 and
recorded in a digital VTR represented by the SD standard. and when
the sync block formats are generated by the recording format
generating means, m transport packets having been input are used to
generate n lines of sync blocks of data. The recording format on
the recording media is configured such that the data of n sync
blocks is disposed on the same track when the data of the sync
block formats obtained by conversion is recorded on the recording
media. Accordingly, the data of the transport packets can be
converted efficiently into sync blocks. Moreover, the n sync blocks
of data is completed within one track, so that when the data of the
sync block format is converted into the data of transport packets
during playback, it is possible to separate into the sets of n sync
block formats using the track information such as the track
identification signal and the sync block number, and the size of
the circuit in the playback system, in particular, can be reduced.
Furthermore, it is not necessary to record the n sync block
identification signal, and the data recording area can be utilized
effectively.
It may be so arranged that when the input intra-picture data is
used at the special playback data generating means to generate the
special playback data, the special playback data is generated in
the form of the input transport packets. The special playback data
generated in the form of the transport packets is, at the time of
recording, converted into sync block format like the transport
packets having been input, and then recorded. Accordingly, when
fast playback is performed by a digital signal playback device, a
circuit for generating transport packets is not required, and it is
possible to share the circuit for converting the sync block format
for playback into transport packets, and the size of the circuit in
a playback-only device, in particular, can be reduced.
It may be so arranged the recording area for the error correction
code is disposed adjacent to the special playback data recording
area. Accordingly, it is possible to reproduce the error correction
code recording area together with the special playback data,
without fail, during fast playback, so that the special playback
data and the error correction code can be efficiently disposed on
the recording media, and it is possible to realize a data
disposition which is immune to the effects of track non-linearity
or the like.
It may be so arranged that when the data having been transmitted in
the transport packets represented by the MPEG2 is recorded in the
digital VTR represented by the SD standard, and the sync block
formats are generated at the recording data generating means, n
lines of sync blocks of data is generated using the m transport
packets. The identification signal for identifying the data
recorded in the error correction code recording area is transmitted
using a recording area provided at the data recording part which
remain when the transport packets are converted into the sync block
formats at the recording data generating means. Accordingly, it is
not necessary to newly provide a recording area for transmitting
the identification signal, and the recording data can be generated
efficiently. Moreover, the recording data for the special playback
data is thereby increased, and the picture quality of-the fast
playback image is improved.
It may be so arranged that when the error correction code is
appended to the output of the recording data generating means, the
output of the recording data generating means and the output of the
part or entirety of the special playback data generating means, to
configure the data block. Interleaving is applied to a collection
of a plurality of data blocks to configure the error correction
code for the normal playback. Accordingly. when the ares for
recording the special playback data are repeated at a period of a
plurality of tracks, and control is so made that the error
correction code is appended only to the normal playback data. it is
not possible to efficiently configure the data blocks and the
maximum burst error correction capability may vary from one track
to another. and the size of the circuit may be increased
unnecessarily, but by forming the data block using part of the
special playback data or the entirety of the data, the freedom in
configuring the data blocks is increased, and for that amount the
size of the circuit for the data block configuration can be
reduced, and the maximum burst error correction capability can be
made uniform between different tracks.
BRIEF DESCRIPTION OF THE DRAWINGS
In the accompanying drawings:
FIG. 1 shows a pattern of tracks in a conventional typical consumer
digital VTR;
FIG. 2 shows a recording format for one track conforming to the SD
standard;
FIG. 3A and FIG. 3B show a data format of the video and audio
signal recording areas on one track for a video signal in
conformity with the SD standard;
FIG. 4 shows the structure of one sync block conforming to the SD
standard;
FIG. 5A and FIG. 5B show the scanning traces of the rotary heads in
the normal playback and fast playback modes in a conventional
digital VTR;
FIG. 6 is a block diagram showing the configuration of the
conventional digital VTR capable of fast playback;
FIG. 7 schematically shows the normal playback and fast playback
modes of the conventional digital VTR;
FIG. 8 shows the scanning trace of a rotary head in typical fast
playback mode;
FIG. 9 is a diagram for explaining overlapping areas in a plurality
of fast playback speeds;
FIG. 10 shows the scanning traces of a rotary head at 5.times. and
9.times. speeds in the conventional digital VTR;
FIG. 11 shows the scanning traces of two rotary heads in 5.times.
speed playback mode of the conventional digital VTR;
FIG. 12 shows the locations of areas on a track in the conventional
digital VTR;
FIG. 13 is a block diagram showing a recording system of a digital
VTR in accordance with Embodiment 1;
FIG. 14 is a block diagram showing a special playback data
generating circuit in Embodiment 1;
FIG. 15 is a block diagram showing a 4.times. speed data generating
circuit in Embodiment 1;
FIG. 16 is a block diagram showing a first error-correction encoder
in Embodiment 1;
FIG. 17A shows a transport packet in an input bit stream;
FIG. 17B shows a record data block (sync block format) to be
recorded on magnetic tape;
FIG. 18A and FIG. 18B show the numbers of sync blocks that can be
reproduced from one track at various fast playback speeds in fast
playback mode;
FIG. 19 shows a pattern of tracks, with a four track period,
including the location of the special playback recording area in
Embodiment 1;
FIG. 20 shows a recording format on magnetic tape in Embodiment
1;
FIG. 21 shows the structure of a data block composed of data for
ten tracks in Embodiment 1;
FIG. 22 is a diagram for explaining conventional interleaving based
on a conventional C4 check code in Embodiment 1;
FIG. 23 is a diagram for explaining interleaving in Embodiment
1;
FIG. 24 is a block diagram showing the configuration of a playback
system for a digital VTR in accordance with Embodiment 1;
FIG. 25 is a flowchart showing an error-correction decoding
algorithm using a C1 check code in Embodiment 1;
FIG. 26 is a flowchart showing a burst error detecting algorithm in
Embodiment 1;
FIG. 27 is a flowchart showing an error-correction decoding
algorithm using a C2 check code in Embodiment 1;
FIG. 28 is a flowchart showing an error-correction decoding
algorithm using the C4 check code in Embodiment 1;
FIG. 29 is a flowchart showing an erasure flag setting algorithm to
be followed for error-correction decoding based on the C4 check
code in Embodiment 1;
FIG. 30 is a flowchart showing an algorithm for setting the maximum
correctable number of erasures, to be followed for error-correction
decoding based on the C4 check code in Embodiment 1;
FIG. 31 is a block diagram showing the configuration of a playback
system for a digital VTR in Embodiment 2;
FIG. 32A to FIG. 32C are diagrams for explaining burst error
detection in Embodiment 2;
FIG. 33 shows an error-correction decoding algorithm using the C1
check code in Embodiment 2;
FIG. 34 shows an error-correction decoding algorithm using the C4
check code in Embodiment 2;
FIG. 35 is a block diagram of a recording system of a digital VTR
in Embodiment 4;
FIG. 36A to FIG. 36C are diagrams showing the arrangement of the
rotary heads of the respective channels on a rotary drum typically
used in a digital VTR of the SD standard;
FIG. 37 shows a track pattern of a period of four tracks including
the disposition of the special playback data recording area in
Embodiment 4;
FIG. 38 shows a recording format on a magnetic tape in Embodiment
4;
FIG. 39A to FIG. 39G show a method of configuring the data blocks
at the time of generating the error correction code appended to the
normal playback data in Embodiment 4;
FIG. 40 shows the method of interleaving over five data blocks
applied to the data when the error correction code is appended to
the normal playback data in Embodiment 4;
FIG. 41 shows the configuration of the error correction code (C5
check code) appended to the special playback data in Embodiment
4;
FIG. 42 is a block diagram showing the playback system of a digital
VTR in Embodiment 4;
FIG. 43 shows the track pattern and the scanning traces of the
rotary heads during 4.times. speed playback using the 1 ch.times.2
drum configuration from a magnetic tape having the recording format
of FIG. 38; and
FIG. 44 shows the track pattern and the scanning traces of the
rotary heads during 18.times. speed playback using the 1 ch.times.2
drum configuration from a magnetic tape having the recording format
of FIG. 38.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
Embodiment 1
FIG. 13 is a block diagram showing the configuration of a recording
system in a digital VTR of Embodiment 1 of the present invention.
In the drawing, reference numeral 1 denotes a transport packet
input terminal. A header analyzing circuit 10 is for detecting a
transport header of a transport packet, also detecting a sequence
header, picture header, and other headers in a bit stream, and
separating intra-frame or -field coded data. A parallel-to-series
converting circuit (hereinafter referred to as P/S converting
circuit) 11 is for performing parallel-to-serial conversion on
input transport packets to produce a bit stream. A special playback
data generating circuit 12 for extracting a bit stream representing
an intra-frame or -field coded picture (which will be referred to
as an intra picture) on the basis of header information detected by
the header analyzing circuit 10, and then producing special
playback data to be reproduced at a 4.times. or 18.times.
speed.
A first memory 13 is for temporarily storing transport packets
received through the input terminal 1 and restructuring data
according to a sync block format shown in FIG. 17B at the time of
output (the details will be described later). A 4.times. speed data
generating circuit 14 uses 4.times. speed playback data supplied
from the special playback data generating circuit 12 to produce
special playback transport packets for 4.times. speed playback. A
18.times. speed data generating circuit 15 uses 18.times. speed
playback data supplied from the special playback data generating
circuit 12 to produce special playback transport packets for
18.times. speed playback.
A second memory 16 is for temporarily storing 4.times. speed
playback data fed in the form of a transport packet and
restructuring data according to a sync block format at the time of
output of the data (See FIG. 17A and FIG. 17B. The details will be
described later). A third memory 17 is for temporarily storing
18.times. playback data fed in the form of a transport packet and
restructuring data according to a sync block format at the time of
output of the data (See FIG. 17A and FIG. 17B. The details will be
described later).
A data synthesizer 18 is for re-arranging the input transport
packets supplied from the first memory 13 and the special playback
data supplied from the second memory 16 and third memory 17 into a
predefined sync block order (the various kinds of data are
restructured according to the sync block format shown in FIG. 17B
in the first memory 13, second memory 16, and third memory 17 and
then fed to the data synthesizer 18). Reference numeral 19 denotes
a fourth memory. A first error-correction encoder 20 is for
producing error correction codes (hereinafter referred to as C4
check codes) to be appended to record data (hereinafter referred to
as record data or record digital data) stored in the fourth memory
19. A fifth memory 21 is for storing record data to which error
correction codes have been appended by the first error-correction
encoder 20. A second error-correction encoder 22 is for appending
error correction codes (C1 check codes and C2 check codes) to the
record data stored in the fifth memory 21 in a horizontal direction
and perpendicular direction respectively that are defined under the
SD standard.
A digital modulator 23 is for digitally modulating record data that
is supplied from the fifth memory 21 with error correction codes
appended. ID information and sync information shown in FIG. 4 are
appended to each sync block data when fed to the digital modulator
23. Reference numeral 24 denotes a recording amplifier, and 25
denotes a rotary drum. A rotary head 26a is for recording or
reproducing data onto or from A tracks. A rotary head 26b is for
recording or reproducing data onto or from B tracks. Reference
numeral 27 denotes magnetic tape.
FIG. 14 is a block diagram showing the special playback data
generating circuit 12 in Embodiment 1. In the drawing, reference
numerals identical to those in the prior art example denote
identical or corresponding elements. Reference numeral 35 denotes
an input terminal through which a bit stream representing an intra
picture that is an intra-frame or -field coded picture is received.
Output terminals 36a and 36b are for outputting 4.times. speed
playback data and 18.times. speed playback data, respectively. A
variable-length decoder 4 is for applying a variable-length
decoding to input intra pictures. Reference numeral 5 denotes a
counter. Data extracting circuits 6a and 6b are for extracting
4.times. speed playback data and 18.times. speed playback data
respectively from the input bit steam of intra data. EOB appending
circuits 7a and 7b are for appending end-of-block (hereinafter EOB)
codes to the end of each DCT block.
FIG. 15 is a block diagram showing the 4.times. speed data
generating circuit 14 in Embodiment 1. The 4.times. speed data
generating circuit 14 and 18.times. speed data generating circuit
15 have the same circuit configuration. In Embodiment 1. detailed
description of the 18.times. speed data generating circuit 15 will
therefore be omitted. In the drawing, reference numeral 40 denotes
an input terminal through which header information representing
transport headers, sequence headers, picture headers and the like,
as well as additional information including a quantization table
which are from the header analyzing circuit 10 are received.
Reference numeral 41 denotes an input terminal through which
4.times. speed playback data supplied from the special playback
data generating circuit 12 is received. A transport header
modifying circuit 42 is for modifying a transport header fed
through the input terminal 40 and outputting a resultant header. A
header appending circuit 43 is for appending header information
representing sequence headers, picture headers and the like
detected by the header analyzing circuit 10 as well as the
additional information (including a quantization table) required
for decoding 4.times. speed playback data, to 4.times. speed
playback data sent from the special playback data generating
circuit 12. A packeting circuit 44 is for performing
serial-to-parallel conversion on the bit stream data output from
the header appending circuit 43 to produce data in which one byte
consists of eight bits, and forming data part of each transport
packet by collecting data of 184 bytes long. A transport header
appending circuit 45 is for appending a transport header sent from
the transport header modifying circuit 42, to the data in a
transport packet produced by the packeting circuit 44.
FIG. 16 is a block diagram showing the first error-correction
encoder 20 in Embodiment 1. In the drawing, reference numeral 50
denotes an input terminal through which a control signal is
received from the data synthesizer 18. An output terminal 51 is for
a memory address output from a shuffling address generator 56. An
output terminal 52 is for a control signal for controlling writing
or reading of data into or from the fourth memory 19 is supplied.
Reference numeral 53 denotes an input terminal through which data
read from the fourth memory 19 is received. Reference numeral 54
denotes an output terminal through which data (C4 check codes) is
supplied to the fourth memory 19. Reference numeral 55 denotes an
output terminal through which a control signal is supplied to the
second error-correction encoder 22. The shuffling address generator
56 produces a shuffling address on the basis of address information
sent from an error-correction coding control circuit 58. An error
correction code generating circuit 57 is for producing third error
correction codes (C4 check codes) using data supplied from the
fourth memory 19. An error-correction coding control circuit 58 is
for controlling the first error-correction encoder 20. Reference
numeral 59 denotes an input terminal through which a coded data
request signal is received from the second error-correction encoder
22.
FIG. 17A and FIG. 17B show a sync block format in Embodiment 1.
FIG. 17A shows a transport packet contained in an input bit stream
(or data). FIG. 17B shows a record data block to be recorded on the
magnetic tape 27. A bit stream received through the input terminal
1 contains a digital video signal, digital audio signal, and
digital data concerning the video signal and audio signal. These
signals and data are transmitted in the form of a transport packet
shown in FIG. 17A. A packet is composed of a header of four bytes
long and a data part of 184 bytes long.
In Embodiment 1, a bit stream is detected in transport packet
units. Two detected transport packets are, as shown in FIG. 17B,
transformed into a record data block (according to a sync block
format) composed of five sync blocks. In the drawing, H1 denotes a
first header. H2 denotes a second header. Recorded in the header H1
is identification data indicating a sync block number of the sync
block of the five sync blocks (For details of one sync block, see
FIG. 4). Recorded in the header H2 is identification data
indicating whether data concerned is video data or audio data,
i.e., and the like.
FIG. 18A shows the numbers of sync blocks that can be reproduced
from one track at different fast playback speeds in fast playback
mode. In the drawings, the numerical values denote the numbers of
sync blocks that can be reproduced from one track at the playback
speeds in special playback mode in which a rotary head of 10 .mu.m
is used (incidentally, a track pitch is defined as 10 .mu.m under
the SD standard). Calculation is performed on the assumption that
the number of sync blocks on one track (corresponding to
180.degree. arc) is 186 and the part of the playback signal whose
output level exceeds -6 dB is acquired, as in the prior art
example.
FIG. 19 shows a track pattern of four track interval, including the
disposition of the special playback data recording areas according
to Embodiment 1. As illustrated, recording areas for a bit stream
(hereinafter referred to as normal playback data), and recording
areas for special playback data and third error correction codes
(C4 check codes) are repeated at an interval of four tracks. Each
group of four tracks will be hereafter called a track format. FIG.
20 shows locations of data in four-track units (data of one track
format data) on the magnetic tape 27, shown in FIG. 19.
Referring to FIG. 18A to FIG. 20, a recording format in Embodiment
1 will be described. Hereinafter, tracks on which data is recorded
by the rotary head 26a will be referred to as A tracks, and tracks
on which data is recorded by the rotary head 26b will be referred
to as B tracks. In FIG. 19, reference numeral T1 denotes a first
track on which data is recorded by the rotary head 26a of channel
A. T2 denotes a second track on which data is recorded by the
rotary head 26b of channel B. T3 denotes a third track on which
data is recorded by the rotary head 26a of channel A. T4 denotes a
fourth track on which data is recorded by the rotary head 26b of
channel B. in Embodiment 1, data is recorded on magnetic tape
according to the aforesaid pattern of four tracks of the first to
fourth tracks; that is, a track format. In FIG. 19, f0, f1, and f2
written under the tracks indicate kinds of pilot signals recorded
on the tracks as reference signals used for tracking control during
playback. In Embodiment 1, the normal playback data, special
playback data, and C4 check codes are recorded in a video area
consisting of 135 sync blocks except a C2 check code recording area
and VAUX data recording areas.
In FIG. 19, reference numerals A0 to A4 denote the locations of
18.times. speed playback data recording areas on the magnetic tape
27. Each of the 18.times. speed playback data recording areas (A0
to A4) is five sync blocks long. Five 18.times. speed playback data
recording areas are, as illustrated, defined on each A track (T1 or
T3). In the drawing, areas bearing the same reference numeral
contain the same data.
Similarly, reference numeral B0 denotes the location of the
4.times. speed playback data recording area on the magnetic tape
27. The 4.times. speed playback data recording area is 25 sync
blocks long and located on the track T2 alone. C0 to C3 denote the
locations of C4 check code recording areas on the magnetic tape 27.
Each C4 check code recording area is ten sync blocks long and
provided on each track.
The number of sync blocks allocated to each data recording area is
determined on the basis of the data shown in FIG. 18A. As apparent
from FIG. 18A, in 4.times. speed playback mode, 62 sync blocks can
be reproduced from each track. In 18.times. speed playback mode,
10.9 sync blocks can be reproduced from each track. FIG. 19 shows
allocation of data for the respective the special playback speeds
onto the magnetic tape 27 on the basis of the data shown in FIG.
18A. Data is recorded on the magnetic tape 27 by repetitively
recording the track format shown in FIG. 19.
FIG. 20 shows a recording format in Embodiment 1. In Embodiment 1,
the information in the area B0 is used to achieve fast playback at
a 4.times. speed. The information in the areas A0 to A4 is used to
achieve fast playback at a 18.times. speed. As for the area B0, the
same special playback data is, as shown in FIG. 20, recorded
repetitively in two track formats. As for each of the areas A0 to
A4, the same data is recorded repetitively in nine track formats.
As for the area B0, the same data is recorded twice, with eight
tracks forming an interval or period. As for each of the areas A0
to A4, the same data is recorded eighteen times, with 36 tracks
forming an interval or period. In the A0 to A4 areas and B0 areas
hatched (or shaded) with the same patterns in FIG. 20, the same
special playback data is recorded.
Next, a method of recording error correction codes which features
the present invention will be described. As was described in
conjunction with the prior art example, an ATV signal (or DVB
signal, or the like) is subjected to data compression based on
motion-compensated prediction. When errors occur in playback data
in normal playback mode, since the playback data is acquired from
an ATV signal, the errors propagate over a plurality of fields or
frames. This results in an impaired view. When a digital VTR
conforming to the SD standard is used as a storage media for
storing data or programs for a computer or the like, some data may
not be reproduced due to drop-outs caused by scratches or dust on
the magnetic tape 27. For restoring the data, it is desired to
apply a more powerful error correction code.
Description will be made of a method of producing C4 check codes to
be recorded in error correction code recording areas. In Embodiment
1, data from ten tracks is collected and interleaved to produce C4
check codes. A block structured by collecting data from ten tracks
in the course of producing C4 check codes will be referred to as a
data block. A (138, 128, 11) Reed-Solomon code is used as the C4
check code. FIG. 21 shows a data block structured using data on ten
tracks for the purpose of interleaving. In the drawing, sync block
numbers written vertically in the drawing are assigned to sync
blocks constituting a video data area in ascending order, starting
from 0 which is assigned to VAUX data at the leading sync block
(sync block address 19) in the video area shown in FIG. 2.
Likewise, data numbers written horizontally are assigned to record
data except the header H1 in the sync block format shown in FIG.
17B. In Embodiment 1, as shown in FIG. 21, C4 check codes are
produced in relation to record data except the ID, sync data, C1
check codes, C2 check codes, and the headers H1 shown in FIG. 17B,
which are recorded in a video data area. However, the present
invention is not limited to such a configuration.
FIG. 22 is an explanatory diagram concerning interleaving to be
performed according to the C4 check code in Embodiment 1. Assuming
that a track number of a track within a data block for producing C4
check codes is Tn (0.ltoreq.Tn.ltoreq.9), a sync block number of a
sync block within the track is SBn (0.ltoreq.SBn.ltoreq.137), and a
data number of data within the sync block is Dn
(0.ltoreq.Dn.ltoreq.75), data D is defined as [Dn, SBn, Tn]. For
interleaving data shown in FIG. 22, the data is arranged as
follows:
(D[0, 0, 0], D[1, 1, 1], D[2, 2, 2], . . . , D[(j mod 76),j,(j mod
10)], . . . , D[50, 126, 6], D[51, 127, 7], . . . , D[60, 136, 6],
D[61, 137, 7])
where 128 bytes of D[0, 0, 0]to D[50, 126, 6] and D[61, 137, 7]
serve as information symbols, and 10 bytes of D[51, 127, 7]to D[60,
136, 6] serve as C4 check codes. FIG. 22 schematically shows the
interleaving. Interleaving is applied to data (symbols) in 138 sync
blocks located along a chain line in FIG. 22. Dotted lines in FIG.
22 indicate directions of interleaving within each track. The C4
check codes are recorded in sync blocks 146 to 155 within a video
area shown in FIG. 2.
The foregoing operation is performed on all data in a leading sync
block of each track. When C4 check codes are produced using data
starting with the i-th data (with data number i) as counted from
the leading sync block, on a k-th track, data is arranged as
follows:
(D[i, 0, k], D[(i+1 mod 76), 1, (k+1 mod 10)], . . . , D[(i+j mod
76), j, (k+j mod 10)], . . . , D[(i+127 mod 76), 127, (k+127 mod
10)], D[(i+128 mod 76), 128, (k+128 mod 10)], . . . , D[(i+137 mod
76), 137, (k+137 mod 10)])
The value i is varied from 0 to 75 for each of ten tracks (the
value k is varied from 0 to 9). Thus, C4 check codes are produced.
In FIG. 22 or the above formulae, (X mod Y) indicates a remainder
obtained when an integer X is divided by an integer Y. C4 check
codes resulting from the above interleaving are recorded in the
predetermined areas shown in FIG. 19.
Next, the burst error correction capability of the conventional C4
check code will be described. As shown in FIG. 22. interleaving and
encoding are applied for a depth of ten tracks. The minimum hamming
distance in the C4 check code is eleven, and the C4 check code can
therefore correct up to ten errors of erasures. In the conventional
interleaving, for example, when attention is directed to a distance
between each pair of interleaved data on a track number 0 shown in
FIG. 22, the distance between symbol D[70, 70, 0] and the next
symbol D[4, 80, 0] is small. That is, while the distance between
the other symbols is ten sync block plus ten symbols, the distance
between the symbols D[70, 70, 0] and D[4, 80, 0] is nine sync
blocks plus ten symbols (actually, eight plus five plus one symbols
for the C1 check code and sync and ID information are added), and
the distance between the symbols D[70, 70, 0]and D[4, 80, 0] is
smaller by about one sync block.
In case a long burst error occurs on one track due, for example, to
drop-outs in normal playback mode, the error-correction capability
of the C4 check code slightly varies depending on the location of
the burst error. When a burst error occurs over the adjoining
symbols D[70, 70, 0] and D[4, 80, 0] that are separated by a small
distance as described above, since the distance is smaller by about
one sync block, a correctable burst error becomes shorter
accordingly (in practice, shorter by about one sync block). In
general, error correction based on the C4 check code is achieved by
correcting erasures according to error-detecting flags set as a
result of error correction based on the C2 check code (hereinafter
referred to as C2 decoding). If the number of errors detected by C2
decoding is not less than a predetermined value, the error
correction based on the C4 check code (hereinafter referred to as
C4 decoding) is executed, disregarding the error-detecting
flags.
If the number of errors detected by C2 decoding is not less than
the predetermined value, C4 decoding is executed, disregarding the
error-detecting flags set as a result of the C2. In this case, the
burst error correction capability of the C4 check code is degraded
to about a half because erasure correction is not performed. With
the conventional interleaving, if a burst error occurs within one
track, the burst error correction capability (a maximum correctable
burst-error length) varies by about one sync block depending on the
location of the burst error or the start position of a codeword.
The same applies to a decoding algorithm according to which erasure
correction is not performed during error correction based on the C4
check code. (The decoding algorithm will be described in detail in
conjunction with the operations of a playback system.)
Next, the interleaving employed in Embodiment 1 will be described
in conjunction with FIG. 23. Like the prior art example, assuming
that a track number of a track within a data block for configuring
C4 check codes is Tn (0.ltoreq.Tn.ltoreq.9), a sync block number of
a sync block within the track is SBn (0.ltoreq.SBn.ltoreq.137), and
data number within the sync block is Dn (0.ltoreq.Dn.ltoreq.75),
data is defined as D[Dn, SBn, Tn]. In this case, the data is
arranged as follows:
(D[0, 0, 0], D[0, 1, 1], D[0, 2, 2], . . . , D[5*INT(j/10) mod 76,
j, (j mod 10)], . . . , D[60, 126, 6], D[60, 127, 7], . . . , D[65,
136, 6], and D[65, 137, 7])
where 128 bytes of D[0, 0, 0] to D[60, 126, 6] and D[65, 137, 7]
serve as information symbols, and ten bytes of D[60, 127, 7] to
D[65, 136, 6] serve as C4 check codes. In the above formulae,
INT(X) means an integer contained in a real number. FIG. 23
schematically shows interleaving. Interleaving is performed on 138
sync blocks of data (symbols) in the direction of the arrow. The C4
check codes are recorded in the same areas as those in the prior
art example.
The foregoing interleaving is performed for all data in a leading
sync block on each track. Assuming that C4 check codes are produced
using data starting with the i-th data in a leading sync block on
the k-th track, the data is arranged as follows:
(D[i, 0, k], D[i, 1, (k+1 mod 10)], . . . , D[(i+5*INT[j/10]) mod
76, j, (k+j mod 10)], . . . , D[(i+60 mod 76), 127, (k+127 mod
10)], D[(i+60 mod 76), 128, (k+128) mod 10)], . . . , D[(i+65 mod
76), 137, (k+137 mod 10)])
The interleaving is executed by varying the value i from 0 to 75
per track, for each of ten tracks (varying the value k from 0 to
9). Thus, interleaving is executed to produce C4 check codes. In
the above formula, (X mod Y) indicates a remainder obtained when an
integer X is divided by an integer Y. INT(X) means an integer
contained in a real number X. Data that has been interleaved to
produce C4 check codes is recorded in predetermined areas shown in
FIG. 19.
Next, the burst error correction capability of the C4 check code in
Embodiment 1 will be described. As shown in FIG. 23, data on ten
tracks is interleaved and encoded. The minimum hamming distance in
the C4 check code is 11, and the C4 check code can therefore
correct errors up to ten erasures. In Embodiment 1, unlike the
prior art example, when attention is directed a distance between
each pair of interleaved data on a track of track number 0 shown in
FIG. 23, the symbols are disposed at equal interval of 10 sync
blocks plus five symbols. In normal playback mode, when a long
drop-out occurs (at whichever position on one track a long burst
error occurs), the error correcting capability of the C4 check code
is exhibited uniformly irrespective of the location of the burst
error.
According to the C4 check code, as shown in FIG. 23, interleaving
is applied to data blocks each consisting of 138 sync blocks to a
depth of ten tracks produce codewords. The minimum hamming distance
in the C4 check code is 11, so that C4 check code can therefore
correct errors up to ten erasures. If no error is detected on other
tracks, the maximum burst error correction capability is 10*10=100
sync blocks. In normal playback mode, even when one hundred sync
blocks of data within one track has not been reproduced because of
drop-outs, for example, the data can be restored by applying the C4
check code.
In a digital VTR conforming to the SD standard. when the frame
frequency is 30 Hz, a digital video signal representing one frame
is recorded on ten tracks. Track numbers are appended to ID signals
of the ten tracks, on which the data of one frame is recorded, in
ascending order from a leading track of the ten track.
Specifically, the same number is assigned to each pair of A and B
tracks, and track numbers ranging from 0 to 4 are used. In the
U.S.A., since the frame frequency is, as well known, 30 Hz, track
numbers are assigned in the above-mentioned manner in a digital VTR
conforming to the SD standard. In Embodiment 1, interleaving is
executed in ten track units, so that it is possible to identify the
data block of ten tracks for which the C4 check code has been
generated, without appending additional information. In Embodiment
1, since the frame frequency is 25 Hz in PAL or SECAM zone such as
Europe, data of one frame is recorded on twelve tracks and track
numbers ranging from 0 to 5 are assigned to the tracks.
Interleaving is therefore executed in twelve track units.
A general expression in accordance with the present invention will
be presented below. Assuming that the number of data in a direction
u in FIG. 23 is n1, the number of effective samples in a direction
w is n3, the number of information symbols of C4 check code is k2,
and the minimum hamming distance in the C4 check code is d3, a
codeword V(Z) is expressed by the following polynomial.
##EQU1##
Herein, .alpha. denotes a parameter determining the length of data
to be interleaved. The value .alpha. is determined so that the
above condition can be satisfied. When the value .alpha. is
determined in such a way that error-detection flags set as a result
of application of the identical C2 check code in one codeword are
not used, a very efficient codeword can be produced. In the above
expression, (X mod Y) means the remainder obtained when an integer
X is divided by an integer Y. INT(X) indicates an integer contained
in a real number X.
Next, the operations of a recording system will be described with
reference to FIG. 13 to FIG. 23. A transport packet received
through the input terminal 1 is fed to the header analyzing circuit
10 and first memory 13. The header analyzing circuit 10 first
detects a transport header in the input transport packet, and then
analyzes the input transport header. The header analyzing circuit
10 then separates a program association table (PAT) and program map
table (PMT) from the transport stream, and then detects a PID
assigned to a program to be recorded in a digital VTR.
The header analyzing circuit 10 separates a transport packet for
transmitting video data of a program to be recorded, on the basis
of the detected PID number. The information of the transport packet
is fed to the first memory 13. The header analyzing circuit 10
analyzes the data in the transport packet, then separates header
information such as sequence headers, picture headers, and slice
headers. On the basis of the header information, the header
analyzing circuit 10 separates intra picture data from the
transport packet. Various header information appended to the intra
picture (hereinafter referred to as intra picture or intra picture
data), and additional information appended to the header
information are also separated.
The sequence header is a header appended to a bit stream of the
video signal. The header contains information for discriminating
between MPEG1 and MPEG2, information indicating an aspect ratio of
a picture, information indicating the transfer rate of picture
data, and the like. The picture header is a header appended to the
start of each frame or field. The picture header points to the
start of each frame or field and contains a mode signal
representing an encoding mode, a quantization table, and the like.
In the MPEG2 system, when one frame of data is transferred, a
screen constructed by one frame (field) is segmented into a
plurality of slices for data transfer. The slice header points to
the start of each slice. (For details of the headers, refer to the
draft on the MPEG2 system).
The header information and associated additional information (for
example, a quantization table) separated by the header analyzing
circuit 10 are supplied to the P/S converting circuit 11, first
memory 13, special playback data generating circuit 12, 4.times.
speed data generating circuit 14, and 18.times. speed data
generating circuit 15. Intra picture data extracted by the header
analyzing circuit 10 is supplied to the P/S converting circuit
11.
The intra picture data (hereinafter referred to as intra-frame
picture data. In the following description, it is assumed that the
intra picture data is coded in one frame unit) detected by the
header analyzing circuit 10 is supplied to the P/S converting
circuit 11, where parallel-to-serial conversion is performed to
produce a bit stream data of one bit. The bit stream data of one
bit which is a serial data having been obtained by the conversion
is fed to the special playback data generating circuit 12.
Referring to FIG. 14, the operations of the special playback data
generating circuit 12 will be described. MPEG2-conforming image
compression is achieved as follows: discrete cosine transform
(hereinafter referred to as DCT) is performed on a video signal in
units of a block of eight lines by eight pixels (hereinafter
referred to as a DCT block); data that has been subjected to DCT
(hereinafter referred to as DCT coefficients) is quantized; and
then DCT coefficients are read sequentially by performing zigzag
scanning, from low-frequency components of the video signal on
which a power spectrum is concentrated, and is subjected to
run-length encoding, wherein the length of runs of "0" coefficients
is encoded (in this way, the video signal is decomposed into
run-length data and coefficient data). The data that has been
run-length encoded is then subjected to two-dimensional
variable-length coding. Thus, this results in a lowered transfer
rate.
The serial data of an intra picture, which has been received
through the input terminal 35, is fed to the variable-length
decoder 4, and data extracting circuits 6a and 6b. The
variable-length decoder 4 performs variable-length decoding on the
input bit stream. In Embodiment 1, the input bit stream is not
perfectly decoded during variable-length decoding, but the run
length of a variable-length codeword and the code length thereof
are detected and supplied. This is intended to reduce the size of
the circuitry. (Needless to say, variable-length decoding may be
accomplished perfectly.) The counter 5 counts the number of DCT
coefficients within each DCT block decoded on the basis of the run
lengths, and supplies the resultant count values to the data
extracting circuits 6a and 6b.
The data extracting circuit 6a extracts variable-length codewords
of 4.times. speed playback data to be transferred (recorded)
according to the predefined code amount control information (the
number of DCT coefficients to be transferred) of 4.times. speed
playback data (in Embodiment 1, 4.times. speed playback data is a
signal to be recorded in area B0, and 18.times. speed playback data
is a signal to be recorded in areas A0 to A4), and the results of
counting sent from the counter 5. An actual data extracting
procedure is as follows: the number of DCT coefficients sent from
the counter 5 is compared with the code value control information;
and then variable-length codewords are then transferred until the
number of DCT coefficients reaches the code value control
information. A demarcation separating variable-length codewords is
detected on the basis of code length information sent from the
variable-length decoder 4.
Similarly, the data extracting circuit 6b extracts variable-length
codewords of 18.times. speed playback data according to code value
control information of 18.times. speed playback data, and
information sent from the counter 5 and variable-length decoder 4.
The EOB appending circuits 7a and 7b then append an EOB code to the
end of each of DCT blocks of the extracted data, and then supply
the resultant data through the output terminals 36a and 36b. The
start of each DCT block is detected by the variable-length decoder
4 and supplied to the counter 5 and the data extracting circuits 6a
and 6b.
The number of extracted DCT coefficients may be the same or
different between different multiple speeds. When the number of DCT
coefficients to be extracted differs between the different multiple
speeds, the number of DCT blocks to be recorded in a special
playback transport packet differs between the multiple speeds. The
number of areas in which special playback data can be recorded is
limited as mentioned above. If the special playback data recording
areas for the respective special playback speeds consist of the
same number of sync blocks, when the number of DCT coefficients to
be recorded in one DCT block increases, a larger number of special
playback data recording areas becomes necessary. This results in a
longer updating period (hereinafter referred to as refresh cycle)
of fast playback picture data in fast playback mode. Incidentally,
a larger number of DCT coefficients results in a higher the
reproduced picture quality. By contrast, a smaller number of DCT
coefficients recorded in one DCT block leads to a smaller amount of
special playback data per frame, and a smaller number of special
playback data recording areas, which in turn result in a shorter
refresh cycle of fast playback picture data. However, reproduced
picture quality deteriorates because of the smaller number of DCT
coefficients. The amount of extracted data for each fast playback
speed should be determined by trade-off between the refresh cycle
and image quality.
The 4.times. speed playback data and 18.times. speed playback data
sent from the special playback data generating circuit 12 are fed
to the 4.times. speed data generating circuit 14 and 18.times.
speed data generating circuit 15, respectively. The subsequent
processing is the same between the different playback speeds
(4.times. speed and 18.times. speed). Herein, description will be
made of production of 4.times. speed playback data. The operations
of the 4.times. speed data generating circuit 14 will be described
with reference to FIG. 15. The 4.times. speed data generating
circuit 14 uses the transport header information and other various
header information (including additional information) sent from the
header analyzing circuit 10 and the 4.times. speed playback data
sent from the special playback data generating circuit 12 to
produce a transport packet for 4.times. speed playback.
The transport header information received through the input
terminal 40 is supplied to the transport header modifying circuit
42, where the transport header is modified. Specifically, the
header information indicating the continuity of transport packets
in the transport headers appended to the transport packets carrying
intra picture data is rewritten according to intra picture
information sent from the header analyzing circuit 10. The header
appending circuit 43 appends in the header information of such as
sequence headers, picture headers, and slice headers, which has
been detected by the header analyzing circuit 10, as well as
information from each header needed to decode special playback data
(for example, an encoding mode flag or quantization table), to a
special playback bit stream received through the input terminal 41
from the special playback data generating circuit 12.
The special playback data with header information appended thereto
is subjected to serial-to-parallel conversion by the packeting
circuit 44 and thus transformed into data in which one byte
consists of eight bits. The resultant data in which one byte
consists of eight bits is segmented into units of 184 bytes,
whereby data part of transport packet is constructed. During the
serial-to-parallel conversion, zeros are inserted in front of each
header information so that each header information will fit in four
bytes in conformity with the MPEG2 system. (Each header information
is 32 bits long. For production of a transport packet, header
information must fit in four bytes long.) To be more specific, when
header information extend over five bytes (occupying only part of
the first and last of the five bytes), zeros are appended in front
of the header information so that the header information fits in
four bytes. Transport header information from the transport header
modifying circuit 42 is appended to transport packet data of 184
bytes configured at the packeting circuit 44, and output. The
header information is read from the transport header modifying
circuit 42 in synchronism with a timing signal sent from the
packeting circuit 44. 4.times. speed playback data produced by the
4.times. speed data generating circuit 14 is supplied to the second
memory 16 in the form of transport packets.
The above description was made of formation of the 4.times. speed
playback data into transport packets. 18.times. speed playback data
is similarly processed. The 18.times. speed playback data from the
special playback data generating circuit 12 is fed to the 18.times.
speed data generating circuit 15. In the 18.times. speed data
generating circuit 15, the header appending circuit 43 appends
headers and additional information according to the header
information sent from the header analyzing circuit 10. The
packeting circuit 44 carries out serial-to-parallel conversion in
the manner described above, to construct data part of transport
packets. The transport header appending circuit 45 appends modified
transport headers sent from the transport header modifying circuit
52 and then supplies the resultant data to the third memory 17 in
the form of transport packets.
The special playback transport packet data sent from the 4.times.
speed data generating circuit 14 and 18.times. speed data
generating circuit 15 are stored in the second memory 16 and third
memory 17, respectively. The second and third memories 16 and 17
store the input data in the form of transport packets in the
storage area and form special playback data of one frame
(field).
The special playback data of one frame is read from the second
memory 16 and third memory 17 in units of two special playback
transport packets, in response to a data request signal sent from
the data synthesizer 18, and then reconstructed into data of five
sync blocks shown in FIG. 7B, and supplied to the data synthesizer
18. At this time, headers H1 and H2 shown in FIG. 17B are
appended.
Transport packets received through the input terminal 1 are input
and stored in the first memory 13. The input data is read from the
first memory 13 in response to a control signal (data request
signal) sent from the data synthesizer 18. At this time, data
having been input in units of a transport packet is reconstructed
into data of five sync blocks shown in FIG. 17B in units of two
transport packets, and then output. In the same way as with the
special playback data, the H1 and H2 header information is appended
to the sync block data when the sync block data is supplied from
the first memory 13.
The data synthesizer 18 synthesizes the data supplied from the
first memory 13, second memory 16, and third memory 17 according to
a recording format. The operations for generating recording format
will be described below. The data synthesizer 18 generates the
recording format on the basis of a reference signal of the servo
system (including tape travel control system and rotary drum
control system) which are not illustrated. Specifically, when the
reference signal is input, the data synthesizer 18 detects the
track number of a track for which a recording format is defined
next. according to information sent from a track number detection
counter provided in the data synthesizer 18. The data synthesizer
18 also identifies the type of the track (which of the tracks T1 to
T4 the track in question is) of which is recorded next, according
to track identification information of the track within one track
format. When detection is completed, the count values of the
counters are incremented by one.
The data synthesizer 18 determines the type of special playback
data and recording areas thereof according to the result of the
identification for discrimination between the tracks T1 to T4. The
number of repetitions of the special playback data for each speed
is checked, and when the predetermined number of repetitions is
reached, the data request signal is issued so that the next special
playback data will be read from a memory containing the
corresponding special playback data.
To be more specific, when 18.times. speed playback data has been
recorded repeatedly eighteen times, a data request signal is sent
to the third memory 17 so that the next 25 sync blocks of special
playback data will be read out. The 25 sync blocks of the 18.times.
speed playback data read from the third memory 17 is temporarily
stored in a 18.times. speed playback data memory incorporated in
the data synthesizer 18. At this time, the the number of
repetitions of the 18.times. speed playback data is reset to "0".
Likewise, when 4.times. speed playback data has been recorded
twice, a data request signal is supplied to the second memory 16,
such that the next 25 sync blocks of special playback data be
output. The 25 sync blocks of 4.times. speed playback data been
read from the second memory 16 is temporarily stored in a 4.times.
speed playback data memory in the data synthesizer 18. At this
time, the number of repetitions of 4.times. speed playback data is
reset to 0. When the number of repetitions is not more than a
predetermined value, special playback data for each playback speed
stored in the data synthesizer 18 is used to produce the record
data. At this time, the number of repetitions of special playback
data is incremented by one.
When checking of the number of repetitions of special playback data
is completed, data locations within one track are specified using a
track identification signal. When the data locations for one track
are specified, bit stream input in sync block units and the special
playback data formed from the bit stream for the respective speeds
are read from the first memory 13 and data synthesizer 18. One
track of record data is thus generated and supplied to the fourth
memory 19.
The record data of one track which has been produced by the data
synthesizer 18 is temporarily stored in the fourth memory 19. In
response to a data synthesis end signal sent from the data
synthesizer 18, the first error-correction encoder 20 sends data
write addresses and write control signals for storing the data for
one track in the fourth memory 19. When the record data for ten
tracks has been stored in the fourth memory 19, the first
error-correction encoder 20 starts producing C4 check codes. In
Embodiment 1, data blocks are configured according to the track
number information. Specifically, tracks from track number 0 to
track number 9 are collected to structure a data block.
The manner of generating C4 check codes will be described below,
with reference to FIG. 16 and FIG. 23. In response to a data
synthesis end signal fed through the input terminal 50, the
error-correction coding control circuit 58 produces data write
addresses and write control signals and writes record data for one
track, which has been supplied from the data synthesizer 18, in the
fourth memory 19. (The details of the manner of production of the
write addresses and write control signals is omitted.) When data
for ten tracks (a data block) has been stored (structured) in the
fourth memory 19, the error-correction coding control circuit 58
supplies data read addresses and read control signals to the fourth
memory 19, and supplies a C4 check code production start signal to
each of the shuffling address generator 56 and error-correction
code generating circuit 57, so that C4 check codes can be produced.
Embodiment 1, the error-correction coding control circuit 58
produces addresses of data before the interleaving (hereinafter
referred to as relative addresses) at the time of producing C4
check codes.
The data read addresses generated by the error-correction coding
control circuit 58 are fed to the shuffling address generator 56.
The shuffling address generator 56 converts the relative addresses
supplied from the error-correction coding control circuit 58 into
absolute addresses so that data read from the fourth memory 19 will
be interleaved as shown in FIG. 23. The absolute addresses provided
by the shuffling address generator 56 are sent to the fourth memory
19 via the output terminal 51.
128 symbols of data is read sequentially from the fourth memory 19
and supplied to the first error-correction coding circuit 20
according to the address information (absolute addresses
information), which have been interleaved addresses, and read
control signals, supplied from the first error correction coding
circuit 20. The data read from the fourth memory 19 is fed to the
error-correction code generating circuit 57 via the input terminal
53. The error-correction code generating circuit 57 produces C4
check codes on the basis of the input data. When the
error-correction code generating circuit 57 completes producing C4
check codes, a C4 check code production end signal is sent to the
error-correction coding control circuit 58. The C4 check codes
produced by the error-correction code generating circuit 57 are
sequentially stored in the fourth memory 19 via the output terminal
54, responsive to control signals sent from the error-correction
coding control circuit 58. In response to the end signal, the
error-correction coding control circuit 58 produces write addresses
(relative addresses) and write control signals which are used to
write the C4 check codes in the fourth memory 19.
The shuffling address generator 56 performs conversion into
absolute addresses so that the write addresses are interleaved. The
C4 check codes are written at predetermined addresses in the fourth
memory 19, according to the write addresses and write control
signals sent from the first error-correction encoder 20. The first
error-correction encoder 20 interleaves all data (symbols) in a
data block shown in FIG. 23 and produces C4 check codes.
When production of C4 check codes using all symbols in a data block
shown in FIG. 23 is completed, the error-correction coding control
circuit 58 sends a C4 check code production end signal to the
shuffling address generator 56 and error-correction code generating
circuit 57. At the same time, it also sends the C4 check code
production end signal to the second error-correction encoder 22 via
the output terminal 55. In response to the C4 check code production
end signal, the second error-correction encoder 22 reads data with
C4 check codes appended thereto, in units of a track. In response
to a coded data request signal from the second error-correction
encoder 22, data read addresses and control signals are provided by
the first error-correction encoder 20. The coded data request
signal is output for each track. When the coded data request signal
is received via the input terminal 59, the error-correction coding
control circuit 58 provides data read addresses and control signals
to the fourth memory 19.
Data for one track read from the fourth memory 19 is temporarily
stored in the fifth memory 21. Error-correction codes in conformity
with the SD standard are produced by the second error-correction
encoder 22 and appended to the record data stored in the fifth
memory 21 (See FIG. 3A). When data for one track has been
structured within the fifth memory 21, the second error-correction
encoder 22 first reads out record data perpendicularly and produces
C2 check codes. The produced C2 check codes are stored in
predetermined addresses in the fifth memory 21. After production of
the C2 check codes is completed, the second error-correction
encoder 22 reads out record data in the recording direction from
the fifth memory 21 and produces C1 check codes. The produced C1
check codes are stored in predetermined addresses in the fifth
memory 21.
When the second error-correction encoder 22 completes production of
C1 check codes, the record data with the C1 and C2 check codes
appended thereto and stored in the fifth memory 21 is read at
predetermined timing responsive to a reference signal sent from a
servo system (including a tape travel control system and rotary
head phase control signal), not shown. Addresses used to read data
from the fifth memory 21 and control signals shall be supplied by
the second error-correction encoder 22 according to the reference
signal. At this time, a track format conforming to the SD standard
is formed. Specifically, a gap of five bytes long is interposed
between sync blocks for appending a sync signal and ID signal. ITI
areas, subcode areas and gaps between data of a predetermined
length are preserved. The resultant data is then output (See FIG. 2
and FIG. 4.) The output of the fifth memory is fed to the digital
modulator 23.
The digital modulator 23 appends a sync signal and ID signal to the
start of each sync block. The ID signal is appended according to
the track number information sent from the data synthesizer 18.
Data with the ID signals appended thereto is digitally modulated
and supplied to the recording amplifier 24. The digital modulation
is applied to the record data, selecting one of modulation patterns
according to the track identification information (indicating one
of T1 to T4) sent from the data synthesizer 18. The data that has
been digitally modulated is amplified and then recorded on the
magnetic tape 27 by means of the rotary heads 26a and 26b.
Next, description will be made of the configuration of a playback
system in a digital VTR which reproduces data structured according
to the aforesaid record format from the magnetic tape 27. FIG. 24
is a block diagram showing the configuration of a playback system
in a digital VTR of Embodiment 1. Reference numerals identical to
those in FIG. 13 denote identical or corresponding components. In
FIG. 24, reference numeral 70 denotes a playback amplifier, and 71
denotes a digital demodulator. An ID detector 72 is for detecting
an ID signal in the playback digital signal (hereinafter referred
to as a playback digital signal and playback digital data) that has
been digitally demodulated. Reference numeral 73 denotes a sixth
memory. A first error-correction decoder 74 is for correcting or
detecting errors which have occurred in the playback digital
signal, according to the C1 check code and C2 check code that are
conforming to the SD standard. Reference numeral 75 denotes a
seventh memory. A second error-correction decoder 76 is for
correcting or detecting errors which have occurred in the playback
digital signal, according to the C4 check code in normal playback
mode. An eighth memory 77 is for storing a playback digital signal
for normal playback. A ninth memory 78 is for storing special
playback data. A switch 79 is for selecting either of the outputs
of the eighth memory 77 and ninth memory 78 according to a select
signal sent from a playback system control circuit 80.
The playback system control circuit 80 is for informing the first
error-correction decoder 74, second error-correction decoder 76,
and switch 79 of the playback mode set in the digital VTR according
to the mode signal sent via the input terminal 81, and for
outputting the result of detection of the ID signal from the ID
detector 72 to the first error-correction decoder 74 and second
error-correction decoder 76. Reference numeral 81 denotes an input
terminal through which a mode signal is received, and 82 denotes an
output terminal.
Prior to describing the operations of the playback system,
error-correction decoding algorithms followed by the first and
second error-correction decoders 74 and 76 will be described with
reference to FIG. 25 to FIG. 30. FIG. 25 shows an error-correction
decoding algorithm based on the C1 check code (hereinafter referred
to as a C1 decoding algorithm) in accordance with Embodiment 1.
FIG. 26 shows a burst error detecting algorithm in accordance with
Embodiment 1. FIG. 27 shows an error-correction decoding algorithm
based on the C2 check code (hereinafter referred to as a C2
decoding algorithm) in accordance with Embodiment 1. FIG. 28 shows
an error-correction decoding algorithm based on the C4 check code
(hereinafter referred to as a C4 decoding algorithm) in accordance
with Embodiment 1. FIG. 29 shows an error-detection flag
(hereinafter referred to as an erasure flag) setting algorithm at
the time of error correction based on the C4 check code in
Embodiment 1. FIG. 30 shows an algorithm for setting a maximum
correctable number of erasures at the time of the error correction
based on the C4 check code in accordance with Embodiment 1.
Description will be made briefly of examples of error-correction
decoding algorithms used according to the three-dimensional error
correction code which is schematically shown in FIG. 3A and FIG.
23. FIG. 25 shows a C1 decoding algorithm (hereinafter referred to
as C1 decoding) using the C1 check code. When data is reproduced,
errors occurring in a playback signal are first corrected to the
limit of the error correction capability of the C1 check code. When
C1 decoding is started, a syndrome is generated using playback
digital data (codewords) demodulated by the digital demodulator 71.
After the generation of the syndrome is completed, the syndrome is
used to calculate the error positions and the error values. If it
is found that the number of errors is not more than four, error
correction is executed. If it is determined that the number of
errors exceeds four, an error-detection flag is set (hereinafter,
an error-detection flag set as a result of application the C1 check
code is referred to as a C1 error-detection flag or a C1 erasure
flag). In Embodiment 1. the minimum hamming distance in the C1
check code is nine, and the C1 check code can therefore correct up
to four errors.
FIG. 26 shows an example of a burst error detection algorithm using
C1 error-detection flags in Embodiment 1. When C1 decoding of data
from one track is completed, burst error detection is carried out
according to the algorithm shown in FIG. 26. The burst error
detection in Embodiment 1 will be described below. In Embodiment 1,
a burst error is detected on the basis of the continuity of C1
error-detection flags set as a result of application of the C1
check code. Specifically, the number of consecutive C1
error-detection flags set within one track is counted. If the
number of consecutive C1 error-detection flags is not less than a
predetermined value, occurrence of a burst error is recognized.
A burst error detecting algorithm shown in FIG. 26 will be
described below. After the C1 decoding, the first error-correction
decoder 74 reads C1 error-detection flags sequentially from the
sync block with the sync block number 0 (SBn=0) (corresponding to
sync block numbers in FIG. 23) and counts the number (b1) of
consecutive C1 error-detection flags that are set. Specifically,
every time a C1 error-detection flag (flc1(SBn)) is found to be "1"
(In Embodiment 1, when an error in data is detected within the sync
block during C1 decoding, the C1 error detecting flag is set to
"1", and the cl error detecting flag is set to "0" when no error is
detected), the count of consecutive C1 error-detection flags, b1,
is incremented by one. The value b1 is then compared with a
predetermined value (nb). If the value b1 is not less than the
predetermined value, occurrence of a burst error is recognized. If
the value b1 is less than the predetermined value, the value SBn is
incremented by one. A C1 error-detection flag in the next sync
block is then read out.
When the C1 error-detection flag (flc1 (SBn)) is "0" (when an error
has been detected), the value b1 is reset to "0". The SBn value is
then incremented by one, and a C1 error-detection flag in the next
sync block is read out. The foregoing sequence is repeated until a
sync block number indicates 137. If the count of consecutive C1
error-detection flags is less than the predetermined value, it is
recognized that no burst error has occurred. Burst error detection
may be performed on data in 149 sync blocks including C2 check
codes. The burst error detection may be carried out concurrently
with C1 decoding at the step of C1 decoding.
Errors that could not be corrected by the C1 check code are
subjected to correction using the C2 check code. In Embodiment 1,
in the error correction using the C2 check code, for the errors
detected based on the C1 check code, erasure correction is
performed, and for the errors undetected by the C1 check code,
error correction is performed. FIG. 27 shows an algorithm of
decoding using the C2 check code. The C2 decoding is carried out
using C1 error-detection flags and the results of burst error
detection. According to the C2 decoding algorithm in Embodiment 1,
when the C2 decoding is started, the results of burst error
detection are checked. Where a burst error has been detected, the
C2 decoding is terminated. This is because if C2 decoding were
performed, the miscorrection might result.
If no burst error is detected, a syndrome is generated using the
playback digital data (codewords) read from the sixth memory 73. At
this time, the C1 error-detection flags set as a result of
application of the C1 check code are also read out in order to
count the number of erasures. (For the C1 error correction flags,
the erroneous bit positions detected during C1 decoding and the
number of errors may be stored in a predetermined register in the
first error-correction decoder 74.) If the number of erasures does
not exceed the error-correction capability of the C2 check code (in
this embodiment, the minimum hamming distance in the C2 check code
is 12, and correction is performed up to eleven erasures), a
modified syndrome is generated using the error syndrome and the
error detection flags set as a result of application of the C1
check code, and errors detected using the C1 check code are then
subjected to erasure correction. Errors undetected by the C1 check
code are corrected to the limit of the error-correction
capability.
If the number of erasures detected using the C1 check code exceeds
the error-correction capability of the C2 check code, the modified
syndrome is not calculated and error correction is carried out to
the limit of the error-correction capability of the C2 check code
(up to five errors are corrected). Such error correction is
possible because it is highly probable that errors detected by the
C1 check code is false erasures (the values recognized as erroneous
by the C1 check code are actually correct). When errors are
detected by the C2 check code, a C2 error-detection flag is set.
When an attempt is made to perform error correction based on the C2
check code on data in which a burst error is detected, since most
of errors detected by the C1 check code when a burst error has
occurred are real erasures (actually incorrect data), the
probability that error correction based on the C2 check code
succeeds is very low, and miscorrection may be effected. In
Embodiment 1, therefore, when a burst error has been detected, C2
decoding is not executed to avoid the miscorrection.
Errors that could not be corrected by the C2 check code are
subjected to error correction based on the C4 check code. In the
error correction using the C4 check code in Embodiment 1, for the
data for which no burst error has been detected, erasure correction
is applied to the errors detected by the C2 check code, and error
correction is applied to the errors undetected by the C2 check
code. As for the data on a track for which a burst error has been
detected, erasure correction is applied to the errors detected by
the C1 check code, and error correction is applied to the errors
undetected by the C1 check code. FIG. 28 and FIG. 29 respectively
show an algorithm of decoding using the C4 check code and an
algorithm for setting error-detection flags. FIG. 30 shows an
algorithm for setting the number of erasures-to-be-corrected used
for C4 decoding.
The algorithm of C4 decoding in Embodiment 1 will be described in
conjunction with FIG. 28 to FIG. 30. Data that has been subjected
to C1 and C2 decodings by the first error-correction decoder 74 is
supplied from the sixth memory 73 and stored in the seventh memory
75. When structuring of a data block of ten tracks shown in FIG. 23
is completed in the seventh memory 75, the second error-correction
decoder 76 starts C4 decoding. In Embodiment 1, the data block is
structured on the basis of track number information contained in ID
signals appended to the starts of sync blocks. The C4 decoding is
executed on the basis of the C1 error-detection flags, C2
error-detection flags, and results of burst error detection.
According to the C4 decoding algorithm in Embodiment 1, when the C4
decoding is started, the results of burst error detection
concerning the data block are checked. When the data on a track for
which a burst error has been detected is decoded, the C1
error-detection flags are used as erasure flags. For a track on
which no burst error has been detected, the C2 error-detection
flags are used as erasure flags.
FIG. 29 shows an error-detection flag setting algorithm in
Embodiment 1. When structuring of a data block of data on ten
tracks is completed in the seventh memory 75, the second
error-correction decoder 76 reads playback digital data (codewords)
from the seventh memory 75. At this time, the C1 error-detection
flags and C2 error-detection flags appended to the codewords, and
burst error detection information are also read out. The read
error-detection flags are selected according to the burst error
detection information. The error-detection flag setting algorithm
for C4 decoding will be described in conjunction with FIG. 29.
First, a sync block number (SBn) is set to 0 at the start of each
codeword. The above information (the C1 error-detection flags, C2
error-detection flags, and results of burst error detection)
concerning the leading codeword is read out. If the codeword is on
a track on which a burst error has been detected, the appended C1
error-detection flag is set as an error-detection flag. For a
codeword on a track on which no burst error has been detected, the
appended C2 error-detection flag is set as an error-detection flag.
Thereafter, the value SBn is incremented by one. The
error-detection flags appended to the next codeword are set in the
same manner. The foregoing sequence is repeated until the value SBn
becomes 137, whereby error-detection flags are set.
Next, an algorithm for setting a maximum correctable number of
erasures during error-correction decoding based on the C4 check
code will be described in conjunction with FIG. 30. Concurrently
with the setting of error-detection flags, the second
error-correction decoder 76 determines the maximum correctable
number of erasures according to the algorithm shown in FIG. 30.
When the C4 decoding is started, first, it is determined whether a
burst error has occurred in the data block (on ten tracks). If no
burst error is detected in the data block, the maximum coffee;table
number of erasures m is set to nine (up to eight erasures can be
corrected).
In Embodiment 1, if no burst error occurs, erasure correction is
not performed to the limit of the error correction capability of
the C4 check code. This is intended to minimize the number of
miscorrections at the time of the C4 decoding because of errors
undetected by the C2 check code during C2 decoding. In particular,
the C4 check code employed in the digital VTR is applied in order
to improve the capability to correct a burst error. In fact, the
minimum hamming distance in the C4 check code is shorter than that
in the C2 check code. When error correction is performed to the
limit of the error correction capability of the C4 check code
during C4 decoding, where the symbol error rate is poor, the result
of the C4 decoding may have a larger number of miscorrections than
the C2 decoding. This phenomenon has been confirmed by computer
simulation. It has been found that a main situation prone to the
miscorrection is that ten errors are detected by the C2 check code
and one error is undetected by the C2 check code. Likewise, a main
situation prone to the miscorrection during C2 decoding is that
eleven errors are detected by the C1 check code and one error is
undetected by the C1 check code. In short, miscorrection is
attributable to miscorrection occurring in the preceding decoding
stage.
In Embodiment 1, the erasure correction capability of the C4 check
code is degraded in order to lower the probability of occurrence of
miscorrection in random error correction. When a burst error is
detected, the maximum correctable number of erasures is set to
eleven. Accordingly, burst error correction can be effected to the
limit of the error correction capability of the C4 check code (up
to ten erasures are corrected). (As mentioned above, a burst error
involving a maximum of 10*10=100 sync blocks per track can be
corrected.)
Taking the above into consideration, the C4 decoding algorithm will
be described in conjunction with FIG. 28. When the C4 decoding is
started, the second error-correction decoder 76 uses playback
digital data (codewords) read from the seventh memory 75 to
generate a syndrome. Concurrently with the generation of the
syndrome, error-detection flags are set according to the algorithm
shown in FIG. 29. The maximum correctable number of erasures is
also determined at the same time. When the generation of the
syndrome is completed, the number of erasures is counted on the
basis of the error-detection flags. If the number of erasures is
less than m (where m is determined according to the algorithm shown
in FIG. 30. In Embodiment 1, if a burst error occurs, error
correction is effected to the limit of the error correction
capability of the C4 check code; that is, up to ten erasures are
corrected. If no burst error is detected, up to eight erasures are
corrected), a modified syndrome is generated using the
above-mentioned syndrome and the error detection flags set by the
above algorithm. Erasure correction is then performed on errors
detected by the C1 and C2 check codes. Errors undetected by the C1
and C2 check codes are corrected to the limit of the error
correction capability.
If the number of erasures detected by the C1 and C2 check codes is
not less than m, no modified syndrome is generated, but error
correction is effected to the limit of the error correction
capability of the C4 check code (up to five errors are corrected).
The error correction is possible because it is highly probable that
errors detected by C1 and C2 check codes are false erasures (values
are recognized as erroneous by the C1 and C2 check codes, but are
actually correct). When an error is detected by applying the C4
check code, a C4 error-detection flag is set.
Next, the normal playback operations of a playback system for a
digital VTR in accordance with Embodiment 1 will be described in
conjunction with FIG. 24 to FIG. 30. The data reproduced from the
magnetic tape 27 by means of the rotary heads 26a and 26b on the
drum 25 is amplified by the playback amplifier 70 and fed to the
digital demodulator 71. The digital demodulator 71 performs data
detection on the input playback signal, converts the data into
playback digital data, and then digitally demodulates the digital
data. ID signal appended to the start of each sync block is
detected by the digital demodulator 71 and supplied to the ID
detector 72. The track number and sync block number are extracted
from the ID signal fed to the ID detector 72, and supplied to the
playback system control circuit 80. The playback system control
circuit 80 supplies an operation mode for the playback system to
the first error-correction decoder 74, second error-correction
decoder 76, and switch 79 according to a mode signal fed via the
input terminal 81. The track number information and sync block
number information sent from the ID detector 72 are supplied to the
first error-correction decoder 74 and second error-correction
decoder 76 via the playback system control circuit 80.
The playback digital data digitally demodulated by the digital
demodulator 71 is stored in the sixth memory 73. In the sixth
memory 73, the playback digital data supplied from the digital
demodulator 71 is accumulated to collect data from one track to
structure an error correction code block shown in FIG. 3A and FIG.
3B. When an error correction code block shown in FIG. 3A and FIG.
3B has been structured, the first error-correction decoder 74 uses
the C1 and C2 check codes to correct or detect errors occurring
during playback, according to the algorithms shown in FIG. 25 to
FIG. 27. Completion of structuring of an error correction code
block is detected on the basis of the sync block number information
and track number information supplied from the playback system
control circuit 80.
The operations of the first error-correction decoder 74 will be
described briefly. When one error correction code block is
structured in the sixth memory 73, the first error-correction
decoder 74 first executes the C1 decoding. Specifically, the
playback digital data is read in units of one sync block from the
sixth memory 73, and the C1 decoding is executed according to the
C1 decoding algorithm shown in FIG. 25. (Up to four errors are
corrected.) Errors detected during the C1 decoding (C1
error-detection flags) are stored at predetermined addresses in the
sixth memory 73. When the C1 decoding is completed, the first
error-correction decoder 74, following the burst error-correction
detecting algorithm shown in FIG. 26, reads error-detection flags
set as a result of C1 decoding from the sixth memory 73, and
recognizes occurrence of a burst error according to the continuity
of the C1 error-detection flags. At this time, the error-detection
flags read from the sixth memory 73 are stored in a register in the
first error-correction decoder 74. The burst error detection may be
executed concurrently with the C1 decoding.
When the reading of the C1 error-detection flags (burst error
detection) is completed, the first error-correction decoder 74
starts the C2 decoding according to the C2 decoding algorithm shown
in FIG. 27. Specifically, as mentioned above, if a burst error is
detected, the C2 decoding is terminated. If no burst error is
detected, first, a syndrome is generated. When the generation of
syndrome is completed, the number of erasures is counted on the
basis of the C1 error-detection flags. If the number of erasures is
less than twelve, a modified syndrome is generated using the error
syndrome and the error detection flags set as a result of
application of the C1 check code, and erasure correction is
performed on errors detected by the C1 check code. Errors
undetected by the C1 check code are also corrected to the limit of
error correction capability.
If the number of erasures detected by applying the C1 check code is
not less than twelve, the modified syndrome is not calculated but
error correction is executed to the limit of the error correction
capability of the C2 check code (up to five errors are corrected).
Errors detected by C2 decoding (C2 error-detection flags) are
stored at predetermined addresses in the sixth memory 73.
The playback digital data that has been subjected to error
correction by the first error-correction decoder 74 is supplied
track by track to the seventh memory 75. The playback digital data
that has been subjected to C1 and C2 decodings, and the C1 and C2
error-detection flags are stored at predetermined addresses in the
seventh memory 75. The results of burst error detection for each
track are supplied from the first error-correction decoder 74 to
the second error-correction decoder 76, and then stored in a
predetermined register in the second error-correction decoder 76.
At this time, the second error-correction decoder 76 determines,
according to the algorithm shown in FIG. 30, whether a burst error
has occurred within the ten tracks, and sets a maximum correctable
number of erasures for the C4 decoding.
The operations of the second error-correction decoder 76 will be
described briefly. When the first error-correction decoder 74
completes the C1 and C2 decodings for one track, the C2 decoding
end signal is sent to the second error-correction decoder 76. At
this time, the burst error detection information is also supplied.
In response to the C2 decoding end signal, the second
error-correction decoder 76 checks the track number signal sent
from the playback system control circuit 80. Based on the track
number information, the second error-correction decoder 76 supplies
write addresses, which are used for writing data read from the
sixth memory 73 at predetermined addresses in the seventh memory
75, and control signals. Reading of data from the sixth memory 73
is performed on the basis of read addresses and control signals
supplied from the first error-correction decoder 74.
When data from the ten tracks is structured in the seventh memory
75, the second error-correction decoder 76 starts error correction
based on the C4 check code. In the seventh memory 75, first,
interleaving is performed in the same way as the one performed
during recording (interleaving shown in FIG. 23). The playback
digital data is then read and fed to the second error-correction
decoder 76. At this time, the data read addresses and control
signals are supplied from the second error-correction decoder 76.
When the playback digital data (codewords) has been read from the
seventh memory 75, the second error-correction decoder 76 first
generates a syndrome according to the algorithm described in FIG.
28. At this time, the C1 error-detection flags and C2
error-detection flags appended to the playback digital data
(codewords) are also read from the seventh memory 75. Erasure flags
(error-detection flags) for the C4 decoding are set according to
the algorithm shown in FIG. 29.
When the generation of syndrome and setting of erasure flags are
completed, the second error-correction decoder 76 counts the number
of erasures. If the number of erasures is less than m, the erasure
flags are used to calculate the modified syndrome and erasure
correction is executed. At this time, errors undetected by the C1
and C2 decodings are also corrected. If the number of erasures is m
or more. error correction is executed with the erasure flags
ignored. The error-detection flags set as a result of the C4
decoding (C4 error-detection flags) are stored at predetermined
addresses in the seventh memory 75.
When the second error-correction decoder 76 completes the C4
decoding of data from the ten tracks, the playback digital data is
read from the seventh memory 75, and written in the eighth memory
77 and ninth memory 78. At this time, the special playback data
(4.times. speed playback data and 18.times. speed playback data)
reproduced from the special playback data recording areas is stored
in the ninth memory 78, while the playback digital data for normal
playback is stored in the eighth memory 77. At this time, the C4
check codes are removed. Reading of data from the seventh memory 75
and writing of data into the eighth and ninth memories 77 and 78
are carried out on the basis of addresses and control signals sent
from the second error-correction decoder 76.
As for the playback data stored in units of a sync block (sync
block format) shown in FIG. 17B in the eighth memory 77, playback
data from five sync blocks are collected according to headers H1
and H2 during reading of playback data. The headers H1 and H2 are
then removed from the data, wherein two transport packets are
restored and supplied. The playback data transformed into the two
transport packets is supplied to the switch 79.
The switch 79 selects the output of the eighth memory 77 during
normal playback, according to the select information sent from the
playback system control circuit 80. The header information H1 and
H2 is, as described above, removed from normal playback data stored
in the eighth memory according to the sync block format shown in
FIG. 17B at the time of data reading, wherein transport packets are
restored (See FIG. 17A) and sent to the switch 79. The normal
playback data read from the eighth memory 77 is supplied through
the output terminal 82 by way of the switch 79.
Next, fast playback operations will be described. The data
intermittently reproduced from the magnetic tape 27 by means of the
rotary heads 26a and 26b on the rotary drum 25 is amplified by the
playback amplifier 70 and fed to the digital demodulator 71. The
digital demodulator 71 detects data in the input playback data,
converts the data into playback digital data, and then digitally
demodulates the digital data. The ID signal appended to the start
of each sync block is detected by the digital demodulator 71. The
ID signals detected by the digital demodulator 71 are fed to the ID
detector 72. The ID detector 72 detects the track number
information and sync block number information in the ID signals and
sends the results of detection to the playback system control
circuit 80.
The playback system control circuit 80 sends a mode signal (an
identification signal indicating normal playback or fast playback
(18.times. speed, 4.times. speed, -16.times. speed, or -2.times.
speed playback)) received through the input terminal 81 to the
first error-correction decoder 74, second error-correction decoder
76, and switch 79. The track number information and sync block
number information supplied from the ID detector 72 are sent to the
first error-correction decoder 74 and second error-correction
decoder 76.
The playback digital data that has been digitally demodulated by
the digital demodulator 71 is stored in the sixth memory 73. In
fast playback mode, for example, as shown in FIG. 8, data is
intermittently reproduced from tracks by means of the rotary head
26a and 26b. It is therefore impossible to collect data for one
track and to structure an error-detecting code block shown in FIG.
3A and FIG. 3B, or FIG. 21 for performing the C2 or C4 decoding. In
Embodiment 1, error correction based on the C2 check code and C4
check code is not executed in fast playback mode.
When the playback digital data is input to the sixth memory 73, the
recording areas of the desired fast playback data are separated on
the basis of the fast playback mode signal (18.times. speed,
4.times. speed, -16.times. speed, or -2.times. speed) and ID
signals (track numbers and sync block numbers) sent from the
playback system control circuit 80, and only the special playback
data is stored temporarily in the sixth memory 73.
The special playback data stored in the sixth memory 73 is read in
units of a sync block, and is subjected to the error correction
based on the C1 check code (C1 decoding) at the first
error-correction decoder 74 to correct or detect errors occurring
during fast playback. When the C1 decoding of data for one sync
block is completed, the first error-correction decoder 74 sends the
C1 decoding end signal to the second error-correction decoder 76,
and the playback digital data that has been subjected to the error
correction by the first error-correction decoder 74 is read
sequentially in units of a sync block from the sixth memory 73 and
supplied to the seventh memory 75. Control signals and addresses in
the sixth memory 73 from which the digital data is read and
supplied from the first error-correction decoder 74. Control
signals and addresses in the seventh memory 75 into which playback
digital data is written are supplied from the second
error-correction decoder 76 in response to the C1 decoding end
signal sent from the first error-correction decoder 74.
As described above, in the fast playback mode, the error correction
based on the C4 check code is not executed, and special playback
data stored in the seventh memory 75 is read and fed to the ninth
memory 78, according to data read addresses and control signals
sent from the second error-correction decoder 76. The output of the
seventh memory 75 is also fed to the eighth memory 77, but, in the
fast playback mode, the playback digital data is not written.
The second error-correction decoder 76 supplies write addresses and
control signals, for storing the special playback data that has
been reproduced according to track numbers and sync block numbers
extracted from ID information, at the predefined addresses in the
ninth memory 78. As for the special playback data stored in the
ninth memory 78 according to the sync block format shown in FIG.
17B, the playback data for the five sync blocks are collected
according to headers H1 and H2 during data reading, and the headers
H1 and H2 are removed, and the original, two transport packets are
restored and output. The special playback data transformed into two
transport packets is then sent to the switch 79.
In fast playback mode, the switch 79 selects the output of the
ninth memory 78 according to the select information sent from the
playback system control circuit 80. As for normal playback data
stored in the ninth memory 78 according to the sync block format
shown in FIG. 17B, the header information H1 and H2 is removed
during data reading, and the original transport packets are
restored (See FIG. 17A) and sent to the switch 79. The special
playback data supplied from the ninth memory 78 is sent through the
output terminal 82 by way of the switch 79.
Embodiment 2
FIG. 31 is a block diagram showing the configuration of a playback
system for a digital VTR in accordance with Embodiment 2. In the
drawing, reference numerals identical to those in FIG. 24 denote
identical or corresponding circuits. Reference numeral 90 denotes
an envelope detector for detecting the waveform of the playback
signal. 91 denotes a burst error detector for detecting a burst
error from the playback signal.
FIG. 32A to FIG. 32C diagrams for explaining the burst error
detection performed by the burst error detector 91 in Embodiment 2.
FIG. 33 shows an error-correction decoding algorithm using the C1
check code in Embodiment 2. FIG. 34 shows an error-correction
decoding algorithm using the C4 check code in Embodiment 2.
The error-correction decoding algorithms based on the
three-dimensional error correction code in accordance with
Embodiment 2, which are shown in FIG. 3A and FIG. 23, will be
described briefly. The burst error detection in accordance with
Embodiment 2 will be described in detail. When data has been
reproduced, first, as in Embodiment 1, errors occurring in the
playback signal are corrected using the C1 check code to the limit
of the error correction capability of the C1 check code. When the
C1 decoding is started, the playback digital data (codewords)
provided by the digital demodulator 71 generates the syndrome. When
the generation of syndrome is completed, the syndrome is used to
detect error positions and error values. If it is found, as a
result of the detection of the error positions and error values
that the number of errors is not more than four, error correction
is executed. If it is found that the number of errors exceeds four,
a C1 error-detection flag (flc1) is set. In Embodiment 2, the flag
(field) is also set with regard to to data on which four errors are
corrected during the C1 decoding (the flag flc1d is set to "1" when
four errors are corrected, and set to "0" in other situations).
Errors that have not been corrected by the C1 check code are
subjected to error correction based on the C2 check code. During
error correction based on the C2 check code in Embodiment 1, errors
detected by the C1 check code are subjected to erasure correction,
and while errors undetected by the C1 check code are subjected to
error correction. The C2 decoding algorithm employed in this
embodiment is similar to that in Embodiment 1. During error
correction of the playback digital data in a data block shown in
FIG. 23 by the C2 check code, a burst error may be detected within
the data block. In this case, in the C2 decoding for the subsequent
playback digital data, in addition to the error-detection flags set
as a result of C1 decoding, the error-detection flags appended to
data in sync blocks in which four errors have been corrected (sync
block flags flc1d are set to one) are set to "1" to indicate that
decoding should be executed, taking that errors have been detected.
The reason for using the error-detection flags set as a result of
the C1 decoding is to minimize the number of undetected errors
during the C1 decoding. Using the newly set error-detection flags,
the C2 decoding is performed on the subsequent playback digital
data within the above-mentioned data block according to the
decoding algorithm shown in FIG. 27.
In Embodiment 2, when a burst error is detected in a data block, as
mentioned above, the subsequent C1 error detection flags are added
to the C1 error detection flag, as mentioned above, such that the
C2 decoding is applied to the sync blocks for which four errors
have been corrected. Accordingly, it is possible to reduce
undetected errors during C2 decoding, and the miscorrection during
C4 decoding is reduced. With regard to tracks on which a burst
error has been detected, C1 error-detection flags are set as
mentioned above, so that, in burst error correction, the number of
errors undetected by the C1 decoding can be minimized.
Consequently, miscorrection during the C4 decoding can be
reduced.
It may be so arranged that, when a burst error is detected in a
data block during C2 decoding, the error correction capability of
the C2 check code is degraded during the subsequent C2 decoding.
This is also effective in minimizing the number of errors
undetected by the C2 decoding, and reducing the miscorrection
during C4 decoding. Specifically, the C2 decoding algorithm may be
changed such that when C2 decoding is performed on the playback
digital data of a data block in which a burst error has been
detected, the maximum correctable number of erasures is decreased,
for example, from eleven to nine, and the maximum correctable
number of erasures for the case of the erasure number being 10 or
more is decreased from five to four, and yet similar effects are
obtained.
For errors that have not been corrected by the C2 check code. the
error correction using the C4 check code is executed. In the error
correction based on the C4 check code in Embodiment 2, erasure
flags are set for executing the erasure correction for the errors
detected by the C2 check code, with regard to the data in the error
correction block for which no burst error is detected. As for data
on a track on which a burst error has been detected,
error-detection flags are set for executing erasure correction for
errors detected by the C1 check code. (The C1 error-detection flags
include the aforesaid flags flc1 as well as flags flc1d appended to
data in sync blocks to which error correction has been applied.)
The error-detection flag setting algorithm is the same as the one
in Embodiment 1 (See FIG. 29), and the detailed description of the
algorithm is therefore omitted. (The C1 error-detection flags are
handled as explained previously.)
Next, a C4 decoding algorithm in accordance with this embodiment
will be described in conjunction with FIG. 34. A data block is
structured as shown in FIG. 23, in the same way as in Embodiment 1,
and the detailed description will therefore be omitted. The C4
decoding is executed, in the same way as in Embodiment 1, on the
basis of the C1 error-detection flags, C2 error-detection flags,
and results of burst error detection. In Embodiment 2, as in
Embodiment 1, when the C4 decoding is started, the results of burst
error detection concerning the data block are checked. For decoding
data on a track on which a burst error has been detected, the C1
error-detection flags are used as erasure flags. As for data on a
track on which no burst error has been detected, the C2
error-detection flags are used as erasure flags.
In Embodiment 2, as in Embodiment 1, when no burst error has
occurred, erasure correction is not effected to the limit of the
error correction capability of the C4 check code. This makes it
possible to minimize the number of miscorrections by the C4 check
code because of errors undetected by the C2 decoding.
The C4 decoding algorithm in Embodiment 2 will be described in
conjunction with FIG. 34. When the C4 decoding is started, the
second error-correction decoder 76 uses the playback digital data
(codewords) read from the seventh memory 75 to generate the
syndrome. Concurrently with the generation of the syndrome, setting
of the error-detection flags is carried out as described previously
(See FIG. 29). At this time, it is checked if a burst error has
occurred in a data block shown in FIG. 23. If a burst error is
detected in the data block, only the erasure correction is
executed, and errors undetected by the C1 or C2 decoding are not
corrected. In this way, miscorrection by the C4 decoding at the
time of occurrence of the burst error is reduced.
When no burst error is detected in the data block, the
error-detection flags are checked to count the number of erasures,
upon completion of the generation of syndrome. If the number of
erasures is less than nine, the generated syndrome is used to work
out a modified syndrome, and erasure correction is then executed
for errors detected by the C1 and C2 check code. At this time,
errors undetected by the C1 and C2 check codes are also corrected
to the limit of the error correction capability.
If the number of erasures detected by the C1 and C2 check codes is
nine or more. no modified syndrome is worked out, but error
correction is executed to the limit of the error correction
capability of the C4 check code (up to five errors are corrected).
The error correction is possible because it is highly probable that
errors detected by the C1 and C2 check codes are false erasures
(values which are recognized as erroneous by the C1 and C2 check
codes, but are actually correct). When errors are detected by
applying the C4 check code, C4 error-detection flags are
output.
Next, the normal playback operations of the playback system for a
digital VTR in accordance with Embodiment 2 will be described in
conjunction with FIG. 31 to FIG. 34. The data reproduced from the
magnetic tape 27 by means of the rotary heads 26a and 26b on the
drum 25 is amplified by the playback amplifier 70 and fed to the
digital demodulator 71. The digital modulator 71 performs data
detection on the input playback data, converts the data into
playback digital data, and then performs digital demodulation. The
ID signal appended to the start of each sync block is detected by
the digital demodulator 71 and supplied to the ID detector 72. The
track number and sync block number in the ID signal fed to the ID
detector 72 are detected and the ID signal supplied to the playback
system control circuit 80. On the basis of a mode signal received
through the input terminal 81, the playback system control circuit
80 supplies an operation mode for the playback system to the first
error-detection decoder 74, second error-detection decoder 76, and
switch 79. The track number information and sync block number
information sent from the ID detector 72 are supplied to the first
error-detection decoder 74 and second error-detection decoder 76
via the playback system control circuit 80.
The output of the playback amplifier 70 is also fed to the envelope
detector 90. The operations of the envelope detector 90 and burst
error detector 91 will be described in conjunction with FIG. 32A to
FIG. 32C. FIG. 32A shows the waveform of a playback signal sent
from the playback amplifier 70. FIG. 32B shows the waveform of an
output of the envelope detector 90 after the playback envelope. The
envelope (See FIG. 32B) of the playback signal sent from the
playback amplifier 70 (See FIG. 32A) is detected by the envelope
detector 90. The output of the envelope detector 90 is fed to the
burst error detector 91. The burst error detector 91 compares the
envelope detected by the envelope detector 90 with a predetermined
level (See FIG. 32B). If the period for which the playback output
continues to be below the predetermined level exceeds a
predetermined period, a burst error is recognized. The results of
burst error detection thus performed (See FIG. 32C) are fed to the
first error-detection decoder 74.
The playback digital data digitally demodulated by the digital
demodulator 71 is fed to the sixth memory 73. In the sixth memory
73, the playback digital data supplied from the digital demodulator
71 for one track are collected and then an error correction code
block shown in FIG. 3A and FIG. 3B is structured. When the
structuring of an error correction code block shown in FIG. 3A and
FIG. 3B is completed, the first error-correction decoder 74
corrects or detects errors, which have occurred during playback,
using the C1 check code and C2 check code according to the
algorithms shown in FIG. 33 and FIG. 27. (The C1 decoding and C2
decoding are identical to those in Embodiment 1, so that their
description is omitted. The four-error correction flags are stored
at predetermined addresses in the sixth memory 73. During the C2
decoding, the C1 error-detection flags and four-error correction
flags are read simultaneously, and, depending on whether or not a
burst error has occurred in a data block, the C1 error-detection
flags are treated as described previously.) During the C2 decoding,
whether or not a burst error has occurred is determined on the
basis of the results of burst error detection provided by the burst
error detector 91.
The playback digital data having been subjected to error correction
by the first error-correction decoder 74 is sent track by track to
the seventh memory 75. The seventh memory 75 stores the playback
digital data that has been subjected to the C1 and C2 decodings, C1
and C2 error-detection flags, and four-error correction flags set
as a result of the C1 decoding, at predetermined addresses. At this
time, the result of burst error detection for each track is
supplied from the first error-correction decoder 74 to the second
error-correction decoder 76 and stored in a predetermined register
in the second error-correction decoder 76.
Next, the operations of the second error-correction decoder 76 will
be described briefly. The description of the operations that are
the same as those in Embodiment 1 will be partially omitted. When
the C1 and C2 decodings on data for one track are completed, the
first error-correction decoder 74 sends the C2 decoding end signal
to the second error-correction decoder 76. At this time, the burst
error detection information is also output. When the C2 decoding
end signal is received, the second error-correction decoder 76
checks the track number signal sent from the playback system
control circuit 80. On the basis of the track number information,
the writing addresses required for writing data read from the sixth
memory 73 at predetermined addresses in the seventh memory 75, and
control signals are supplied.
When data for ten tracks is structured in the seventh memory 75,
the second error-correction decoder 76 starts error correction
based on the C4 check code. First, interleaving is carried out In
the seventh memory 75 in the same manner as the one (shown in FIG.
23) performed during recording. Thus, the playback digital data is
read and fed to the second error-correction decoder 76. At this
time, the data read addresses and control signals are supplied from
the second error-correction decoder 76. When the playback digital
data (codewords) is read from the seventh memory 75, the second
error-correction decoder 76 generates a syndrome according to the
algorithm shown in FIG. 34. The C1 error-detection flags, C2
error-detection flags, and four-error correction flags set as
result of C1 decoding, which correspond to the playback digital
data (codewords) are also read from the seventh memory 75. Erasure
flags (error-detection flags) are then set as a result of the C4
decoding according to the algorithm shown in FIG. 29.
When the generation of the syndrome and setting of erasure flags
are completed, the second error-correction decoder 76 checks if a
burst error has occurred in a data block. If a burst error is found
in the data block, only the erasure correction is executed on the
basis of the error-detection flags set according to the algorithm
shown in FIG. 29. At this time, error correction is not performed
on the undetected errors at the time of the C1 and C2 decodings.
(If there are errors undetected by the C1 and C2 decodings, the
erroneous codewords (the playback digital data) are subjected to
error detection.) If no burst error is found, the number of
erasures is counted. If the number of erasures is less than nine,
erasure flags are used to generate a modified syndrome, and erasure
correction is then executed. At this time, the errors undetected by
the C1 and C2 decodings are corrected. If the number of erasures is
nine or more, the erasure flags are ignored and error correction is
executed. Error-detection flags (C4 error-detection flags) set as a
result of the C4 decoding are stored at predetermined addresses in
the seventh memory 75.
When the second error-correction decoder 76 completes the C4
decoding of data for ten tracks, the playback digital data is read
from the seventh memory 75 and sent to the eighth memory 77 and
ninth memory 78. At this time, the special playback data (4.times.
speed playback data and 18.times. speed playback data) reproduced
from the special playback data recording areas is fed to the ninth
memory 78, while the normal playback playback digital data is fed
to the eighth memory 77. At this time, the C4 check codes are
removed. When the playback data stored in the eighth memory 77
according to the sync block format shown in FIG. 17B is read out,
the playback data for five sync blocks is collected according to
the headers H1 and H2. The headers H1 and H2 are then removed,
whereby the original, two transport packets are restored and
supplied. The playback data transformed into transport packets is
sent to the switch 79.
The switch 79 selects the output of the eighth memory 77 in the
normal playback mode according to the select information sent from
the playback system control circuit 80. The headers H1 and H2 are
removed from the normal playback data, which has been stored
according to the sync block format shown in FIG. 17B, during the
data reading, whereby the original transport packets are restored
(See FIG. 17A) and sent to the switch 79. The normal playback data
read from the eighth memory 77 is sent through the output terminal
82 by way of the switch 79.
In Embodiment 1 or Embodiment 2, the C1, C2, and C4 decoding is
executed as described above, so that even if a burst error occurs,
the number of undetected errors can be restrained adequately and
deterioration of picture quality in the normal playback mode can be
reduced. Furthermore, a burst error can be detected by the
aforesaid configuration, and error-correction decoding can be
achieved successfully. This results in excellent picture quality in
the normal playback mode. When the digital VTR is used as a storage
media for a computer or the like, a burst error involving up to one
hundred sync blocks can be corrected, and the digital VTR can be
used to record programs or the like. Where the sixth and seventh
memories 73 and 75 are used in common for error-correction
decoding, when the C1, C2 and C4 decoding is performed after data
is temporarily stored in the memory and a data block is structured,
it may be so arranged that erasure correction is performed on the
data to which four-error correction flag is appended when C2 and C4
decoding is performed after the detection of a burst error in the
data block. This is effective in reducing the number of errors
undetected by the C2 and C4 decoding. Excellent reproduced pictures
can be obtained.
Embodiment 3
Since the digital VTR's in Embodiment 1 and Embodiment 2 have the
aforesaid configurations, the playback data rate for special
playback data for different fast playback speeds can be set to
large values, and this results in the improved picture quality in
the fast playback mode, and in the normal playback, miscorrection
during the C4 decoding can be reduced adequately, and the normal
playback can therefore be achieved successfully.
In Embodiment 1 and Embodiment 2, the format shown in FIG. 19 (or
FIG. 20) is adopted as a recording format. The recording format is
not limited to the illustrated one, but may be any other recording
format by which the special playback data is separated from the
input data, and recorded in predefined areas on a recording medium.
In a digital signal recording apparatus, playback device, or
recording/playback device (for example, a digital VTR or digital
disk player) that adopts such a recording format, the number of
repetitions of special playback data in the same recording format
is switched depending on the recording mode, such that the special
playback data can be recorded efficiently, and, in addition, the
data rate of the special playback data in the fast playback mode
can be improved and the picture quality in the fast playback mode
can be upgraded. The decoding algorithms shown in FIG. 25 to FIG.
30, FIG. 33, and FIG. 34 may be used for decoding digital data
having been three-dimensionally coded and transmitted.
In Embodiment 1 and Embodiment 2, a (85, 77, 9) Reed-Solomon code
is applied as an error correction code in a recording direction,
and a (149, 138, 12) Reed-Solomon code is applied in a
perpendicular direction, and error-correcting code blocks on ten
tracks are collected to structure a data block. A (138, 128, 11)
Reed-Solomon code is applied as a C4 check code. Thus, a
three-dimensional error correction code is configured.
Alternatively, a (149, 139, 11) Reed-Solomon code or a (135, 125,
10) Reed-Solomon code (the code is structured using the video data
except VAUX) may be used as the C4 check code. The error correction
codes for error correction in the recording direction and
perpendicular direction are not limited to the above examples.
Furthermore, the error correction codes are not limited to
Reed-Solomon codes but may be BCH codes or the like. Nevertheless,
the aforesaid effects can be exerted.
For the error-correction decoding algorithms described in FIG. 25
to FIG. 30, FIG. 33, and FIG. 34, the shuffling pattern for the C4
check codes is not limited to the one shown in FIG. 23. For
example, the shuffling pattern shown in FIG. 22 may also be used,
and yet similar effects are obtained. The combination of the
error-correction decoding algorithms shown in FIG. 25 to FIG. 30,
FIG. 33, and FIG. 34 is not limited to the examples in Embodiment 1
and Embodiment 2. The procedure of detecting a burst error is not
limited to the examples in Embodiment 1 and Embodiment 2. Burst
error detection information obtained through another procedure may
be used to control the C1, C2, and C4 decoding, and the aforesaid
effects can still be exerted. The C1, C2, and C4 decoding
algorithms are not limited to the ones in Embodiment 1 and
Embodiment 2.
The C2 decoding algorithm shown in FIG. 27 need not be used in
combination with the C4 decoding algorithm but can still enables
successful decoding. The C2 decoding algorithm is not limited to
the one shown in FIG. 27, but where the C4 decoding is used to
correct all burst errors, but, according to a conceivable error
correction algorithm, part of the error correction capability of
the C2 check codes is allocated to the error detection during C2
decoding, and the probability of occurrence of errors undetected by
the C2 decoding (miscorrection) can be reduced. (For example, when
the maximum correctable number of erasures is nine, if the number
of erasures is not less than ten, the number of errors to be
corrected is determined to be four.)
The C4 decoding algorithm shown in FIG. 28 or FIG. 34 need not be
combined with the C2 decoding algorithm, but still enables
successful decoding. The C4 decoding algorithm is not limited to
the one described in FIG. 28 or FIG. 34. Alternatively, when the C2
decoding algorithm is used to reduce the number of undetected
errors, the error correction may be performed to the limit of the
error correction capability of the C4 check code during the C4
decoding. With this alternative, similar advantages are attained.
(For example, the maximum correctable number of erasures may be set
to 10 without regard to the presence or absence of the burst error
detection, or the maximum correctable number of erasures may be set
to 9 when the burst error is not detected.) According to another C4
decoding algorithm, if a burst error is detected, only the erasure
correction is executed but errors undetected by the C1 and C2 check
codes are not corrected. This algorithm has the similar
advantages.
The record data is not limited to an ATV signal or DVB signal. In
Japan where a video signal is compressed in conformity with the
MPEG2 system, the present invention can apply to recording of an
ISDB signal or a signal compressed according to the MPEG1 system.
The fast playback speeds are not limited to the 4.times. speed and
18.times. speed. Alternatively, the special playback data recording
areas and fast playback speeds may be defined and determined
according to the playback speeds required for a digital signal
recording/playback apparatus. Even in this alternative, as long as
identical track format is adopted for recording input data, the
aforesaid effects are exerted.
For recording data, which has been transmitted in the form of
transport packets typified by those conforming to the MPEG2 system,
in a digital VTR typified by those conforming to the SD standard,
according to Embodiment 1, two transport packets are converted into
five sync blocks, and recorded. The present invention is not
limited to this. When forming the sync block format, m transport
formats may be used to generate data of n lines of sync blocks. (m
and n denote positive integers.) By configuring the recording
format on the recording media such that, when recording data of the
sync block format obtained by the conversion, data of n sync blocks
are arranged on the same track, the data of the transport packets
can be efficiently converted into the sync block format. Since all
the data of n sync blocks is recorded on the same track, when the
sync block data is transformed into transport packets during
playback, groups of n sync block formats can be separated easily
using the track information and a sync block number, of a track
identification signal, or the like. In particular, the circuitry of
a playback system can be scaled down. It is unnecessary to record
identification signals of n sync blocks, and the data recording
areas can be used effectively. Moreover, the length of one sync
block is not restricted to the one shown in FIG. 4.
The locations of the 4.times. speed playback data recording areas,
18.times. speed playback data recording areas, and error-correction
code recording areas, or the numbers of the areas are not limited
to those described. The track period need not be four tracks. In
Embodiment 1, the fast playback speeds are set to 4.times. and
18.times. speeds as an example. The present invention is not
limited to these speeds but may apply to other multiple speeds. As
long as special playback data recording areas are defined along the
scanning traces of the rotary heads 26a and 26b as described
previously, the similar effects can be obtained.
Embodiment 4
FIG. 35 is a block diagram showing the configuration of a recording
system in a digital VTR of another embodiment (Embodiment 4) of the
invention. Reference numerals identical to those in FIG. 13 denote
identical or corresponding members.
A third error correction encoder 118 is for appending special
playback error correction code (hereinafter referred to as C5 check
code) to respective special playback data output from the second
memory 16 and third memory 17. A second data synthesizer 119 is for
rearranging the input transport packets output from the first
memory 13 and the data output from the third error correction
encoder 118, into the predefined order of sync blocks. An input
terminal 120 is for input of commands and servo information.
An error correction code control circuit 121 is for judging whether
the recording mode of the error correction code during recording is
a special playback error correction code recording mode or a normal
playback error correction recording mode, and outputting a control
signal indicating which of the error correction codes (special
playback error correction code and the normal playback error
correction code) is to be recorded in an error correction code
recording area, to be described later, to the third error
correction encoder 118, the second data synthesizer 119, the second
error correction code circuit 123, a fifth error correction encoder
124 and the digital modulating circuit 23.
Reference numeral 122 denotes a tenth memory, 123 denotes a fourth
error correction encoder for appending error correction code
(hereinafter referred to as C4 check code) to normal playback data,
and 124 denotes a fifth error correction encoder for appending
error correction codes in horizontal (recording) direction (C1
check code) and vertical (perpendicular) direction (C2 check code),
defined by the SD standard, to the data output from the tenth
memory 122. The digital modulator 23 applies digital modulation to
the data output from the fifth error correction encoder 124. ID
information and sync information are appended to the data of each
sync block at the time of input to the digital modulator 23.
The configuration of the special playback data generating circuit
12 in Embodiment 4 is similar to that shown in FIG. 14.
The configurations of the 4.times. speed data generating circuit 14
and the 18.times. speed data generating circuit 15 in Embodiment 4
is similar to that shown in FIG. 15.
FIG. 36A to FIG. 36C are diagrams illustrating the arrangement of
the rotary heads of the respective channels. FIG. 36A to FIG. 36C
show the typical arrangements of the rotary heads 26a and 26b on
the rotary drum 25 used with the SD standard. The system with the
rotary head 26 having the arrangement of FIG. 36A or FIG. 36B will
hereinafter called a 9000 rpm system. The rotary head arrangement
shown in FIG. 36A is represented by "1 CH.times.2", the rotary head
arrangement shown in FIG. 36B is represented by "2 CH.times.1", and
rotary head arrangement shown in FIG. 36C is represented by "2
CH.times.2". The system with the rotary head 26 shown in FIG. 36C
will hereinafter be called a 4500 rpm system. The following
description relates to the system shown in FIG. 36A.
The recording format within a track used in this embodiment is also
as shown in FIG. 2. The data packet used in this embodiment is as
shown in FIG. 17A and FIG. 17B.
In Embodiment 4, a bit stream is detected in transport packet
units. Two detected transport packets are, as shown in FIG. 17B,
transformed into a record data block (sync block format) composed
of five sync blocks. In the drawing, reference numeral H1 denotes a
first header. H2 denotes a second header. Recorded in the header H1
is identification data indicating a sync block number of the sync
block of the five sync blocks (data areas within one sync block is
composed of 77 bytes of data as shown in FIG. 4). Recorded in the
header H2 is identification data indicating whether data concerned
is video data or audio data, i.e., and the like. The sync byte
appended to the head of each transport header need not be recorded.
It is assumed in connection with Embodiment 4 that all the data
within the transport packet is recorded.
The numbers of sync blocks that can be reproduced from one track at
different fast playback speeds in fast playback mode are as shown
in FIG. 18B.
FIG. 37 shows a track pattern of a period of four tracks, including
disposition of the special playback data of Embodiment 4. As
illustrated, special playback data recording areas and error
correction code recording areas on the tracks are repeated at an
interval of four tracks. The four tracks will hereinafter referred
to as a track format. FIG. 38 shows disposition of the data at an
an interval of four tracks (data of one track format), as shown in
FIG. 37, on the magnetic tape.
Referring to FIG. 18B, FIG. 37 and FIG. 38, a recording format in
Embodiment 4 will be described. Hereinafter, tracks on which data
is recorded by the rotary head 26a will be referred to as A tracks,
and tracks on which data is recorded by the rotary head 26b will be
referred to as B tracks. In FIG. 37, reference numeral T1 denotes a
first track on which data is recorded by the rotary head 26a of
channel A. T2 denotes a second track on which data is recorded by
the rotary head 26b of channel B. T3 denotes a third track on which
data is recorded by the rotary head 26a of channel A. T4 denotes a
fourth track on which data is recorded by the rotary head 26b of
channel B. In Embodiment 4, data is recorded on magnetic tape
according to the aforesaid pattern of four tracks of the first to
fourth tracks; that is, a track format. In FIG. 37, f0, f1, and f2
written under the tracks indicate kinds of pilot signals recorded
on the tracks as reference signals used for tracking control during
playback. In Embodiment 4, the normal playback data, special
playback data, and C4 check codes are recorded in a video area
consisting of 135 sync blocks except a C2 check code recording area
and VAUX data recording areas. In FIG. 37, the video area of 135
sync blocks other than the C2 check code recording area and VAUX
data recording area in the video area is shown.
In FIG. 37, A0 to A4 denotes the disposition of the 18.times. speed
playback data recording areas on the magnetic tape. The width of
each of the 18.times. speed playback data areas (A0 to A4) is 5
sync blocks, and the 18.times. speed playback data areas are at
five locations on each of the A tracks (T1 and T3), as illustrated.
The areas with identical reference marks are used for recording
identical data.
In the same way, B0 denotes the disposition of a 4.times. speed
playback data recording area on the magnetic tape. The width of the
4.times. speed playback data recording area is 25 sync blocks. The
4.times. speed playback data recording area is provided at one
location on T2 track, as illustrated.
In the same way, D0 to D5 denote the disposition of the error
correction code recording areas on the magnetic tape. The width of
each of the error correction code recording areas (D0 to D5) is 5
sync blocks. The recording area of D1 is provided adjacent the
4.times. speed data recording area. The area of D4 is provided at
the same height (position in the longitudinal direction of the
track) as the recording area of D1.
The number of sync blocks allocated to each data recording area is
determined by the data shown in FIG. 18B. That is, it is seen from
FIG. 18B that, with a 9000 rpm system, 62 sync blocks can be
obtained from one track during 4.times. speed playback. During
18.times. speed playback, 10.9 sync blocks can be obtained from one
track. On the basis of these, the data disposition on the magnetic
tape for the respective special playback speeds is as shown in FIG.
37.
The input transport packets are recorded in the video area other
than the special playback recording areas (A0 to A4, B0), error
correction code recording areas (D0 to D5), C2 check code and VAUX
recording areas. (The area will hereinafter referred to as a main
area.)
The method of recording error correction codes featuring this
embodiment will next be described. As was described in connection
with the prior art example, data compression is applied to the ATV
signal, using a method of compression based on motion compensated
prediction. When, therefore, an error is detected in the playback
data, the error may propagate a plurality of fields or frames, and
the resultant picture is disagreeable. Where the digital VTR of the
SD standard is used as storage media for storing data or programs
for computers or the like, it has been desired to have a strong
error correction capability to restore data which has not been
reproduced, due to drop-out which occurs because of scratches on
the tape or dust attached on the magnetic tape.
In the fast playback, the error correction using the C1 check code
alone is applied to the playback data, as was described in
connection with the prior art example, so that error occurrence
frequency is high, and the playback picture during special playback
is disagreeable.
The amount of appended data which can be recorded on the magnetic
tape is limited, as was also described in connection with the prior
art example. According to Embodiment 4, therefore, in order to
effectively utilizes the limited data areas, the areas to which the
error correction code is appended, are disposed on traces scanned
by the rotary heads 26a and 26b, and where the user places
importance on the picture quality during normal playback, the data
is generated such that the error correction code is appended to the
normal playback data, and where the user places importance on the
picture quality during special playback, the data is generated such
that the error correction code is appended to the special playback
data.
In this way, the limited data recording areas can be effectively
utilized, and the error correction code can be appended to the
desired data. When the digital VTR is used as a storage media for
computers or the like, identical recording format can be used to
configure powerful error correction code. (The error correction
code for normal playback is written in the error correction code
recording areas.)
The data recorded in the error correction code recording area will
next be described taking account of the above. As shown in FIG. 37,
the recording areas for the error correction code are provided at
six locations in one track format, and each consists of five sync
blocks. In the mode for recording the error correction code
(hereinafter referred to as C4 check code) for normal playback
data, C4 check code is recorded at the six locations.
In the mode for recording error correction code (hereinafter
referred to as C5 check code) for special playback data, C5 check
code for 18.times. speed playback data is recorded in the D0 and D3
areas, and C5 check code for 4.times. speed playback data is
recorded in the D1 area. Data of a predefined value, e.g., a
predefined fixed value (for instance all "0"), is recorded in the
D2, D4 and D5 areas. When the normal playback is made from the
magnetic tape on which error correction code for special playback,
since the data in the areas (D2, D4 and D5) are of a fixed value,
it is possible to improve the error correction capability of the C2
code in error correction decoding during normal playback. (With
regard to the above-mentioned symbols in the C2 decoding, the
predefined fixed value is recorded, so that it is possible to
recognize that there is no error, and the error correction
capability can be improved).
The method of configuring the C4 check code and C5 check code will
be described later. The switching of the error correction code
recorded in the error correction code recording area is made based
on the error correction code recording mode signal output from the
error correction code control circuit 121.
The data in the 4.times. speed playback data recording area (B0)
can be reproduced from one track during one scan of the rotary head
26b in 4.times. speed playback. The data in the 18.times. speed
areas A0 to A4 and the error correction code recording area D0 (or
D3), i.e., these six areas, can be reproduced from six tracks
during one scan of the rotary head 26a in 18.times. speed
playback.
When the error correction code recorded in the error correction
code recording area is the C5 check code, the recording area is
used to apply the error correction to the special playback data
during special playback. When it is the C4 check code, it is
disregarded during the special playback. During normal playback,
when the error correction code recorded in the above-mentioned area
is the C4 check code, the error correction code is used to apply
error correction to the playback digital data. When it is the C5
check code, the data is disregarded during the normal playback.
Details of the method of configuring the error correction code for
the normal playback and the special playback will be described
later.
Details of the playback system will also be described later. But,
according to the data disposition of one track format (recording
format) shown in FIG. 37, the rotary heads 26a and 26b scan the ITI
areas and the sub-code areas on the magnetic tape during 4.times.
speed playback and 18.times. speed playback. That is, the during
special playback, tracking control is achieved using the pilot
signals f0, f1 and f2 read from the ITI areas, and the additional
information, such as time information and item-of-music number
information recorded in the sub-code area, is reproduced.
By repeating the track format shown in FIG. 37, data is recorded on
the magnetic tape. FIG. 38 shows the recording format on the
magnetic tape. In the figure, the 4.times. speed playback data
(recorded in the B0 area) is recorded twice (see FIG. 38), and the
18.times. speed playback data is recorded nine times. That is,
identical 18.times. speed playback data is recorded on two tracks
within one track format (four tracks), so that, in all, identical
data is recorded 18 times. (The special playback data for 18.times.
speed is switched at an interval of 36 tracks.) In FIG. 38, the
areas with identical reference marks are used for recording
identical data.
In FIG. 38, special playback data with an error correction code
appended thereto is recorded in the error correction code recording
area. With regard to the 18.times. speed playback data, a1, a2, a3,
a4 and a5 (a6, a7, a8, a9 and a10) are used for generating the C5
check code. The generated C5 check code is recorded in d18,1
(d18,2) area. That is, data of al, a2, a3, a4, a5, and d18,1 form
one error correction block for 18.times. speed playback data.
Similarly, the data of a6, a7, a8, a9, a10, and d18,2 form one
error correction block for 18.times. speed playback.
With regard to the 4.times. speed playback data, the data in the
4.times. speed playback data recording area and the error
correction code recording area provided adjacent the 4.times. speed
playback data recording area is used to form one error correction
block for 4.times. speed playback data. Specifically, b1 and d4,1,
or b2 and d4,2, or b5 and d4,5 are used to form one error
correction block for 4.times. speed playback data. A fixed value
data is recorded in the area denoted by d0,0 in the mode for
recording error correction code for special playback. In the mode
for recording error correction code for normal playback, the C4
check code is recorded in the error correction code recording areas
(D0 to D5).
Next, FIG. 39A to FIG. 39G, FIG. 40 and FIG. 41 are used for
describing the code configuration of the error correction code used
in normal playback and special playback. First, the code
configuration of the error correction code (C4 check code) for
normal playback will be described. In Embodiment 4, data of 10
tracks are collected, and the collected data is interleaved to
generate C4 check code. (245, 230, 16) Reed-Solomon code is used as
the C4 check code. An example of interleaving will next be
described with reference to FIG. 39A to FIG. 39G and FIG. 40.
FIG. 39A to FIG. 39G are diagrams for explaining the method of
configuring the two-track data blocks for configuring the error
correction code appended to the normal playback data according to
this embodiment. FIG. 39A and FIG. 39B show a specific example of
disposition of the ATV data, special playback data and error
correction code (C4 check code or C5 check code) on A and B tracks.
In the drawing, "A" and "B" respectively denote data of A and B
tracks. In Embodiment 4, to effect interleaving over 10 tracks,
first adjacent two tracks (T1 and T2 tracks, or T3, T4 tracks are
made to form a pair) are used to form a two-track data block of 245
sync blocks. (see FIG. 39G)
The method of configuring a two-track data block of 245 sync blocks
will next be described. In configuring data of 245 sync blocks, the
18.times. speed playback data recorded on the A tracks are removed.
The reason for this is as follows. The error correction code
adopted in the SD standard is formed of elements of Galois field
defined by GF(2.sup.8). Accordingly, the maximum code length of the
Reed-Solomon code defined by the elements of the Galois field is
255. If all the data on two tracks were used, the code length would
be 270 and the error correction code could not be configured.
According to this embodiment, the 18.times. speed playback data is
removed to configure the two-track data block for normal playback.
With regard to the 4.times. speed playback data, data is recorded
only on T2 track, as shown in FIG. 37, so that normal playback
error correction code is appended. This is to avoid increase in the
size of the circuit for the identification of T2 and T4, and for
configuring the two-track data block. In Embodiment 4, the 4.times.
speed playback data is treated in the same way as normal playback
data so as to reduce the size of the circuit. When the 4.times.
speed playback data is removed, it is necessary to configure two
two-track data blocks using the data of 4 tracks when configuring
the two-track data block which will be described later, so that the
control will be complicated and the circuit size will be
increased.
In the digital VTR of the SD standard, when the frame frequency is
60 Hz, one frame of digital video signal is recorded on 10 tracks.
The track number is appended in the ID signal in sequence from the
head of the 10 tracks for recording one frame of data.
Specifically, an identical number is appended to a pair of A and B
tracks, so that track numbers No. 0 to No. 4 are appended. In the
U.S.A., the frame frequency is 30 Hz, so that, in the digital VTR
of the SD standard, track numbers are appended in the above
described manner. In Embodiment 4, interleaving is effected over 10
tracks, and C4 check code can be appended to the data without
appending additional information. When Embodiment 4 is applied to
the PAL/SECAM zone, such as Europe, the frame frequency is 25 Hz,
one frame of data is recorded on 12 tracks, so that track numbers 0
to 5 are appended. The interleaving is therefore effected at an
interval of 12 tracks.
If the elements of the Galois field for the error correction code
is changed to GF(2.sup.9), the generator polynomial for generating
the error correction code and the bit width of one symbol will
vary, so that they cannot be used in common with the error
correction code circuit (e.g., multipliers, dividers, on the Galois
field) conforming to the SD standard, and the bit width of one
symbol is increased, so that the size of the circuit is enlarged.
Accordingly, in Embodiment 4, a (245, 230, 16) Reed-Solomon code is
used as the C4 check code as was described above. During generation
of the error correction code for normal playback, the 18.times.
speed playback data is removed to form a two-track data block of
245 sync blocks. With regard to the 4.times. speed playback data,
although not in use during normal playback, this area is present
only on T2 track, and if this area were removed, it would be
necessary to configure the two blocks from four tracks, and the
track period for generating the C4 check code would be be 20
tracks, so that the size of the circuit would be enlarged.
According to Embodiment 4, therefore, the 4.times. speed playback
data is used for generating the error correction code for normal
playback. In Embodiment 4, the depth of interleaving is 10
tracks.
The method of configuring the error correction code according to
Embodiment 4 will next be made, taking account of the above. First,
the method of configuring a two-track data block of 245 sync blocks
will be described with reference to FIG. 39A to FIG. 39G. In FIG.
39A, the 18.times. speed playback data areas [13] to [17] are first
removed, as described above. The error correction code recording
area [10] is also removed from the data area at the time of
generating the C4 check code. FIG. 39C shows the normal playback
data on A track having been extracted. FIG. 39B shows the
disposition of the ATV data, special playback data, and error
correction code data on the B tracks. The area [8] is a recording
area for 4.times. speed playback data on T2 track, but a recording
area for normal playback data on T4 track.
In Embodiment 4, when a two-track data block of 245 sync blocks is
generated, a group of data from the area [7] and area [9], other
than the area [8], (see FIG. 39D) and a group of data from the area
[8] (see FIG. 39E) are generated. The data from the error
correction code recording areas [10], [11] and [12] on A and B
tracks are collected to form a group of data shown in FIG. 39F.
The groups of data shown in FIG. 39C, FIG. 39D and FIG. 39E,
configured in the above described manner, are respectively divided
into 21, 20 and 5 blocks, each consisting of 5 sync blocks. The
data consisting of five sync blocks are rearranged such that, the
data of B'1 block shown in FIG. 39E is at the head, and the data
shown in FIG. 39C and FIG. 39D are disposed alternately, such as in
the order of A1, B1, A2, B2, . . . in the region Gcd. After the A21
data, B'2, B'3, B'4 and B'5 data are disposed, and then error
correction code recording areas [11], [10] and [12] are disposed in
sequence, as shown in FIG. 39G, to form a two-track data block of
245 sync block. Areas with identical reference marks in the
drawings denote identical areas. For instance, the area [1] in FIG.
39A and the area [1] in FIG. 39C are identical. The reference
numerals "0" and "76" indicating the position in the horizontal
direction in FIG. 39G denote positions (addresses) of the data
within the sync block, and the reference numerals "0" and "244"
indicating the position in the vertical direction in FIG. 39G
denote sync block addresses within the two-track data block of 245
sync blocks.
Five two-track data blocks of data, each of the two-track data
blocks consisting of 245 sync blocks and configured in the above
described manner, are collected, and interleaving over 10 tracks is
applied to the normal playback data to generate the C4 check code.
FIG. 40 will next be used to describe the pattern of interleaving
applied to generate the C4 check code.
FIG. 40 is a diagram for explaining the method of interleaving of
five blocks applied to the data applied to the data when appending
the error correction code to the normal playback data according
this embodiment. If the data with the block number of the two-track
data block shown in FIG. 39G represented by Bn
(0.ltoreq.Bn.ltoreq.4), the sync block number of the sync block
within the two-track data block represented by SBn
(0.ltoreq.SBn.ltoreq.244), and the data number of the data within
the sync block represented by Dn (0.ltoreq.Dn.ltoreq.76) is defined
as D [Bn, SBn, Dn], then according to an embodiment of shuffling
shown in FIG. 40,
(D[0, 0, 0], D[1, 1, 1], D[2, 2, 2], . . . , D[(i mod 5), i, (i mod
77)], . . . , D[4, 229, 75], D[0, 230, 76], D[1, 231, 0], . . . ,
D[4, 244, 13])
Here, the 230 bytes of D[0, 0, 0]to D[4, 229, 75] are information
symbols, and 15 bytes of D[0, 230, 76] to D[4, 244, 13] are C4
check code. FIG. 40 schematically illustrates the shuffling. The
interleaving is made in the direction of the chain lines, and
interleaving of five two-track data blocks is effected. (In effect
interleaving of 10 tracks is applied, since the two-track data
blocks are configured as shown in FIG. 39G.) The dotted lines shown
the direction of interleaving within the two-track data block. (One
data is sampled every five two-track data blocks since interleaving
is at an interval of five two-track data blocks.)
This operation is effected for all the data within the sync block
at the head of each two-track data block. That is, when the C4
check code is generated, starting from the k-th data as counted
from the sync block at the head of the j-th two-track data
block,
(D[j, 0, k], D[(j+1 mod 5), 1, (k+1 mod 77)], . . . , D[(j+i mod
5), i (k+i mod 77)], . . . , D[(j+229 mod 5), 229, (k+229 mod 77)],
D[(j+230 mod 5), 230, (k+230 mod 77) , . . . , D[(j+244 mod 5),
244, (k+244 mod 77)])
By varying k from 0 to 76 per two-track data block, and applying
this to five two-track data blocks (varying j from 0 to 4), the
interleaving is achieved to generate the C4 check code. In FIG. 40
and in the above expression, (X mod Y) represents the remainder
obtained when an integer X is divided by an integer Y. The data for
which the C4 check code has been generated through interleaving is
recorded in the predefined areas shown in FIG. 39A and FIG. 39B.
Where the mode for appending the error correction code is a mode
for recording the special playback data error correction code, the
C5 check code is recorded in the error correction code recording
area, and the C4 check code is discarded.
The burst error correction capability of the C4 check code will
next be described. For the two-track data block of 245 sync blocks
shown in FIG. 39G, interleaving of 5 two-track data block deep as
shown in FIG. 40 is applied. Since the minimum hamming distance of
the C4 check code is 16, up to 15 erasures can be corrected. Since
interleaving of five two-track data blocks deep as shown in FIG.
39G is applied, the maximum burst error correction capability is
15.times.5=75 sync blocks. Accordingly, since, in the case of A
tracks, no C4 check code is appended to the 18.times. speed
playback data area, a maximum burst error correction capability is
75+20 (=95) sync blocks. Similarly, for data on B tracks, a maximum
burst error correction capability is 75 sync blocks. With regard to
T2 tracks, 4.times. speed playback data is recorded twice, so that,
in the event of burst error occurrence, with the data in the above
area, substituting identical 4.times. speed data recorded on a
different track (track other than the track on which a burst error
has occurred), the maximum burst error correction capability is
75+25 (=100) sync blocks. It is therefore possible to restore data
by means of the C4 check code, even when 75 sync blocks of data is
not reproduced due to drop-out during normal playback.
Description will next be made of the configuration of the error
correction code appended to the special playback data with
reference to FIG. 41. FIG. 41 shows the configuration of the error
correction code (C5 check code) appended to the special playback
data. (The data configuration shown in FIG. 41 will hereinafter
referred to as error correction block.) A (30, 25, 6) Reed-Solomon
code is used as the C5 check code. FIG. 39A shows the disposition
of the 18.times. speed playback data and error correction code
recorded on the A track. The data of the error correction block
shown in FIG. 41 is partitioned into parts each consisting of five
sync blocks, which are respectively recorded in the areas [13] to
[17] shown in FIG. 39A. The C5 check code is recorded in the area
[10].
The 4.times. speed playback data recorded on B tracks are recorded
in the special playback area (area [8]), 25 sync blocks by 25 sync
blocks.
In Embodiment 4, interleaving with the depth of data being 10
tracks is applied to generate the C4 check code, as described
above. During such operation, the first memory 13, the second
memory 16 and the third memory 17 convert the transport packets
(FIG. 17A) input via the input terminal 1, into data in block units
shown in FIG. 17B. Using the 4.times. speed playback data output
from the third error correction encoder 118 and the ATV signal
output from the first memory 13, and obtained by conversion into
sync block units, a block data of 245 sync blocks shown in FIG. 39G
is generated by means of the second data synthesizer 119. The
18.times. speed playback data output from the third error
correction encoder 118 and the C5 check code appended to the
special playback data for the respective playback speeds are stored
in predefined addresses in a memory within the second data
synthesizer 119.
The two-track data block of 245 sync blocks configured at the
second data synthesizer 119 is stored in a predefined address in
the tenth memory 122. The 18.times. speed playback data and the C5
check code appended to the special playback data for the respective
playback speeds having been stored in the memory in the second data
synthesizer 119 are also read from the second data synthesizer 119
at the same timing as the two-track data block, and stored in
predefined addresses in the tenth memory 122.
The operation of the recording system will next be described with
reference to FIG. 35 to FIG. 41. The transport packets input via
the input terminal 1 are input to the first memory 13 and the
header analyzing circuit 10. The header analyzing circuit 10
detects the transport header from the input transport packets and
separates the transport packets transmitting video data. It then
analyzes the data within the transport packets transmitting the
video data that have been separated, and detects header information
such as a sequence header, a picture header, and a slice header,
and separates intra-pictures from the transport packets. The header
information described above, and additional information appended to
the header information are also separated.
The sequence header is header information provided in the bit
stream of video signal, and includes identification of MPEG1 or
MPEG2. The picture header is appended to the head of each frame or
field, and indicates the start of the frame or field, and includes
a mode signal indicative of the coding mode or the like. In MPEG2,
when one frame (field) of data is transmitted, a picture of one
frame (field) is divided into a plurality of slices. The slice
header indicates the start of the slice. Details of the respective
headers are described in the Draft of the MPEG2, which is hereby
incorporated by reference.
The header information and the additional information (e.g.,
quantization table information) appended thereto, and intra-picture
information detected by the header analyzing circuit 10 are
supplied to the P/S converting circuit 11, the 4.times. speed data
generating circuit 14 and the 18.times. speed data generating
circuit 15.
The P/S converting circuit 11 applies P/S conversion to the data of
the input transport packets, to obtain bit stream data of one bit.
The transport packet data having been converted into one-bit serial
data is input to the special playback data generating circuit 12.
The operation of the special playback data generating circuit 12 is
similar to that described with reference to FIG. 14 in connection
with Embodiment 1.
The operation of the error correction code control circuit 121 will
next be described. In accordance with the command information input
via the input terminal 120, the error correction code control
circuit 121 outputs a signal indicative of the error correction
code recording mode (special playback error correction code
recording mode or normal playback error correction code recording
mode) to the third error correction encoder 118, the second data
synthesizer 119, the fourth error correction encoder 123, the fifth
error correction encoder 124 and the digital modulator 23. It also
outputs a track identification signal of a four-track period and
track identification signal of a ten-track period on the basis of
the servo information output from the travel control and drum
rotation control systems. In Embodiment 4, the switching of the
error correction code is effected by the output of the tenth memory
122.
The special playback transport packet data output from the 4.times.
speed data generating circuit 14 and the 18.times. speed data
generating circuit 15 are stored in the second memory 16 and the
third memory 17. The second memory 16 and the third memory 17 store
the input data in the form of transport packets and form the
special playback data of one frame (field). One frame of special
playback data formed at the second memory 16 and third memory 17
are read from the respective memories, transport packet by
transport packet, and converted into data of 5 sync blocks, as
shown in FIG. 17B, and output to the third error correction encoder
118. The reading from the second and third memories 16 and 17 is
effected on the basis of the control signal output from the third
error correction encoder 118. The H1 and H2 header information is
appended at the time of input to the third error correction encoder
118.
The third error correction encoder 118 collects 25 sync blocks of
the special playback data corresponding to each playback speed,
having been converted into sync blocks, as shown in FIG. 17B, and
form error correction block. The C5 check is then generated for and
appended to each of the 4.times. speed playback data and 18.times.
speed playback data. The 4.times. speed playback data and 18.times.
speed playback data to which the C5 check code has been appended
are input to the second data synthesizer 119. The generation of the
C5 check code is effected on the basis of the control signal (data
request signal) output from the second data synthesizer 119.
The ATV signal transport packets input via the input terminal 1 are
input to and stored in the first memory 13. The data stored in the
first memory 13 is read on the basis of the control signal output
from the second data synthesizer 119. Two transport packets of the
data input in transport units are converted into five sync blocks
of data as shown in FIG. 17B, and output.
On the basis of the track identification signals of 4 tracks and 10
tracks output from the error correction code control circuit 121,
the second data synthesizer 119 forms the two-track data block of
245 sync blocks shown in FIG. 39G. The number of repetitions of the
special playback data is counted by an internal counter, and a data
request signal for the special playback data is supplied to the
third error correction encoder 118. The second data synthesizer 119
first reads the normal playback data for two tracks, from the first
memory 13. The data read from the first memory 13 is stored in a
memory provided in the second data synthesizer 119. Similarly, the
special playback data and the C5 check code output from the third
error correction encoder 118 are also stored in a memory provided
in the second data synthesizer 119. The data stored in the second
data synthesizer 119 are rearranged, and two-track data block of
245 sync blocks shown in FIG. 39G is formed and output to the tenth
memory 122. The switching between the 4.times. speed playback data
and the normal playback data on the B track is effected within the
second data synthesizer using the track information of 4 tracks.
The 18.times. speed playback data and the C5 check code for the
special playback data are output from the second data synthesizer
119, following the two-track data block of 245 sync blocks, to the
tenth memory 122.
The data output from the second data synthesizer 119 is written in
predefined addresses of the tenth memory 122. The writing address
and the control signal for the tenth memory 122 are supplied from
the fourth error correction encoder 123.
The operation for generating the C4 check code will next be
described. When the configuration of the 5 two-track data blocks
are completed in the tenth memory 122, the fourth error correction
encoder 123 outputs the reading address for reading the data for
generating the C4 check code and the control signal to the tenth
memory 122. In accordance with the control signal, data is read
from the tenth memory 122. (The data read is interleaved as shown
in FIG. 40. That is, interleaving is applied to the reading
addresses output from the fourth error correction encoder 123.) The
fourth error correction encoder 123 generates the C4 check code on
the basis of the data read from the tenth memory 122. The C4 check
code generated by the fourth error correction encoder 123 is stored
in the C4 check code storage area within the tenth memory 122. When
the C4 check code for the five two-track data blocks is generated,
the data stored in the tenth memory 122 is output to the fifth
error correction encoder 124.
Reading of the data from the tenth memory 122 is effected on the
basis of the read addresses and the control signal output from the
fifth error correction encoder 124. The fifth error correction
encoder 124 reads the data stored in the tenth memory 122, on the
basis of the mode signal and the track identification signal output
from the error correction code control circuit 121, with the sync
blocks being rearranged in the order shown in FIG. 39A and FIG.
39B. The selection between the C4 check code and C5 check code is
effected by switching between the read addresses (C4 check code
read address and the C5 check code read address) output from the
fifth error correction encoder 124 on the mode signal output from
the error correction code control circuit 121. According to
Embodiment 4, in the case of the special playback error correction
code recording mode. when the C5 check code is read from the tenth
memory 122, fixed value data is recorded in D2, D4 and D5 in FIG.
37. This is achieved by storing in advance the fixed value data in
the predefined address in the tenth memory 122.
The data read from the tenth memory 122 is stored in a memory
within the fifth error correction encoder 124, in track units. When
one track of data is configured, the fifth error correction encoder
124 appends, to the above-mentioned data, the error correction code
according to the SD standard shown in FIG. 3A and FIG. 3B. That is,
for the above-mentioned data recorded in the video area, a C2 check
code is first generated and appended in track units. For the data
with the C2 check code having been appended, a C1 check code is
next generated and appended. With regard to audio areas, C3 check
code and C1 check code are similarly appended. In Embodiment 4, the
audio data areas are not used, and filled with dummy data of a
fixed value, to allow for modification or adaptation which may be
required of the format in the future. The data with the error
correction code having been appended according to the SD
specification is read from the fifth error correction encoder 124,
at predefined timings, in track units. At this time, a track format
according to the SD standard is generated. Specifically, a gap of
five bytes is provided for appending a sync signal, and an ID
signal between adjacent sync blocks, and a gap and the like of a
predefined amount are provided for ITI area, sub-code area and
between data, and the above-mentioned data is then output. The
output of the fifth error correction encoder 124 is input to the
digital modulator 23.
The digital modulator 23 first appends the sync signal and the ID
signal to each sync block. At this time, a flag information
indicating the type of the error correction code is appended in the
ID signal with regard to the sync blocks recorded in the error
correction code recording areas. During playback, the flag in the
ID signal is used to identify the type of the error correction
code. The data with the ID signal having been appended is subjected
to digital modulation, and amplified by the recording amplifier 24,
and recorded on the magnetic tape, by means of the rotary heads 26a
and 26b.
Description will next be made of the configuration of the playback
system of the digital VTR for playing back the magnetic tape having
the above-described recording format. FIG. 42 is a block diagram
showing the playback system of the digital VTR according to this
embodiment. The reference numerals identical to those in FIG. 31
and FIG. 35 denote member having identical configuration and
operating in the same way, so that their description is omitted. In
the figure, a third error correction decoder 152 is for performing
correction and detection of errors in the playback signals using
the C1 check code and C2 check code. The third error correction
decoder 152 detects the ID signal in the playback signal and
outputs it to an error correction decoding control circuit 154, to
be described later.
An input terminal 153 is for input of the mode signal (normal
playback or special playback) indicating the mode of playback being
conducted. The error correction decoding control circuit 154 is
responsive to the ID signal and the playback mode signal output
from the third error correction decoder 152, for outputting a
control signal for switching between the error correction decoding
process for special playback and the error correction decoding
process for normal playback, to a fourth error correction decoder
156 and a fifth error correction decoder 157, and also outputting a
switching signal for a switch 159. The error correction decoding
control circuit 154 also detects the track number information, sync
block address information, information on the type of the error
correction code recorded in the error correction code recording
area. 155 denotes a eleventh memory for storing the normal playback
data. The fourth error correction decoder 156 receives the control
signal (the result of the ID information and the playback mode
signal detected at the third error correction decoder 152) output
from the error correction decoding control circuit 154, and if the
C4 check code is recorded in the above-mentioned error correction
code recording area, applies error correction decoding using the C4
check code to the data stored in the eleventh memory 155, and
corrects or detects errors in the playback signal. The fifth error
correction decoder 157 receives the control signal output from the
error correction decoding control circuit 154, and if the C5 check
code is recorded in the above-mentioned error correction code
recording area, applies error correction decoding using the C5
check code to the special playback data. A twelfth memory 158 is
for storing the special playback data output from the fifth error
correction decoder 157. Reference numeral 160 notes an output
terminal.
Before describing the operation of the playback system, the
operation of the 4.times. speed and 18.times. speed playback of the
digital VTR according to Embodiment 4 is described with reference
to FIG. 43 and FIG. 44. FIG. 43 shows the relationship between the
track pattern and the scanning traces of the rotary heads which
result when 4.times. speed playback is effected with a 1 ch.times.2
drum configuration from a magnetic tape having the recording format
shown in FiG. 38. In the figure, the scanning traces denoted by A
are traces by a rotary head 26a of an A channel. Similarly, the
scanning traces denoted by B are traces by a rotary head 26b of a B
channel. As was described above, the 4.times. speed playback data
is recorded on B tracks, and identical data is recorded repeatedly
for 2 track format periods (identical data is recorded at two
recording areas). Accordingly, as shown in FIG. 43, each of the
rotary heads 26a and 26b scans the scanning periods of the rotary
heads 26a and 26b, so that the recorded 4.times. speed playback
data recorded by the B-channel rotary head 26b can be reproduced.
At the same time, the rotary heads 26a and 26b can reproduce data
from the sub-code area, as shown in FIG. 43. The tracking is
achieved at the ITI area. Details of the 4.times. speed playback
using a digital VTR having a 2 ch.times.1 drum configuration is
omitted, but in short the rotary head 26a scans the left side of
the rotary head 26b in the 1 ch.times.2 drum configuration (in FIG.
43, the scanning traces of the rotary head 26a are on the left side
of and adjoining to the scanning traces of the rotary head 26b),
and it is therefore possible to reproduce the 4.times. speed
special playback data recorded by the B-channel rotary head
26b.
FIG. 44 shows the relationship between the track pattern and the
scanning traces of the rotary heads which result when 18.times.
speed playback is effected with a 1 ch.times.2 drum configuration
from a magnetic tape having the recording format shown in FIG. 38.
In the figure, the scanning traces denoted by A are traces by a
rotary head 26a of an A channel. Similarly, the scanning traces
denoted by B are traces by a rotary head 26b of a B channel. As was
described above, the 18.times. speed playback data is recorded on A
tracks, and identical data is recorded repeatedly for 9 track
formats (identical data is recorded at 18 locations). Accordingly,
as was described in connection with the prior art example, the
18.times. speed playback data recorded by the A-channel rotary head
26a can be reproduced. (As shown in FIG. 44, the 18.times. speed
playback data can be reproduced by the A-channel rotary head 26a.)
At the same time, the rotary head 26a can reproduce the signal from
the sub-code area. Moreover, since the ITI area is also scanned, it
is possible to achieve the tracking at the ITI area. Details of the
18.times. speed playback using a digital VTR having a 2 ch.times.1
drum configuration is omitted, as was the 4.times. speed playback,
but in short the rotary head 26b scans the right side of the rotary
head 26a in the 1 ch.times.2 drum configuration (in FIG. 44, the
scanning traces of the rotary head 26b are on the right side of and
adjoining to the scanning traces of the rotary head 26a), and it is
therefore possible to reproduce the 18.times. speed special
playback data recorded by the A-channel rotary head 26b.
The amount of data which can be reproduced from one track during
4.times. speed playback and 18.times. speed playback is as shown in
FIG. 18B. That is, with a 9000 rpm system, it is 62 SB at 4.times.
speed and 10.9 SB at 18.times. speed. The positions on the tracks
from which the data is reproduced at -2.times. speed and -16.times.
speed are identical to those for the 4.times. speed and 18.times.
speed, so that -2.times. playback and -16.times. speed playback can
be realized with the recording format shown in FIG. 38. Details of
the playback from a magnetic tape having the above-mentioned
recording format, with a 4500 rpm system are omitted, but, in
short, the fast playback can be realized by setting the speed at
2.times. (93 SB) and 8.times. (13 SB). During the fast playback, it
is possible to reproduce, apart from the special playback data, the
error correction code recording area (4.times. speed playback data)
or the error correction code recording area (18.times. speed
playback) provided separately on the track. During normal playback,
the rotary heads 26a and 26b reproduce from all the tracks. The
above-mentioned error correction code recording area can be
reproduced during fast playback (4.times. speed, 18.times. speed,
-2.times. speed and -16.times. speed), and during normal playback,
and the same recording area can be shared. (That is, whichever of
the error correction codes is appended, the error correction code
recording area can be reproduced during fast playback and normal
playback.)
In the above description, the tracking during fast playback is
achieved at the ITI area. The invention is not limited to such an
arrangement. For instance, at the 18.times. speed playback, the
tracking phase may be detected at one of the special playback data
recording areas, and used for control. The tracking phase may
alternatively be detected at a plurality of the special playback
data recording areas, and used for control. With regard to the
4.times. speed playback, the tracking phase may be detected at an
adjacent A track by the rotary head 26a, and used for control.
Furthermore, rough adjustment of the tracking phase may be effected
at the ITI area, and fine adjustment may be be effected at the
special playback area. The above-mentioned tracking control method
is effective when there is a track non-linearity, in a playback on
a compatible device.
The operation of the normal playback of the playback system will
next be described. The data reproduced from the magnetic tape via
the rotary heads 26a and 26b on the drum 25 is amplified by the
playback amplifier 70, and input to the digital demodulator 71. The
digital demodulator 71 detects data from the input playback data,
performs conversion into the playback digital data, and performs
digital demodulation. The playback digital data obtained by digital
demodulation at the digital demodulator 71 is input to the third
error correction decoder 152. When the playback digital data is
input, the third error correction decoder 152 separates the ID
signal appended to the head of each sync block. and outputs the
result of the separation to the error correction decoding control
circuit 154. The playback digital data from which the ID signal has
been separated is supplied to the third error correction decoder
152 where correction and detection of errors occurring during
playback are effected using the C1 check code and C2 check code.
The playback digital data having been subjected to error correction
at the third error correction decoder 152 is supplied to the
eleventh memory 155 and the fifth error correction decoder 157. The
special playback data (4.times. speed playback data and 18.times.
speed playback data) reproduced from the special playback data
recording area is input to the fifth error correction decoder 157.
The playback digital data other than the 18.times. speed playback
data is input to the eleventh memory 155. Incidentally, in
Embodiment 4, during normal playback, the error correction decoding
operation using the C5 check code is not performed. (That is, the
special playback data is input to the fifth error correction
decoder 157, but no error correction decoding operation is not
applied.)
The ID signal detected at the third error correction decoder 152 is
input to the error correction decoding control circuit 154. The
error correction decoding control circuit 154 detects the ID
signals separated from the sync blocks which are recorded in the
error correction code recording areas, and identifies the type of
the error correction code using the appended identification flag
data, and also detects, during normal playback, the numbers of the
track and sync block which are being reproduced. It also detects
the current playback mode from the playback mode signal input via
the input terminal 153. The result of the identification of the
error correction code, the track number, the sync block number and
the playback mode information are supplied to the fourth error
correction decoder 156 and the fifth error correction decoder 157,
and the playback mode signal is supplied to the switch 159.
The playback digital data input in track units to the eleventh
memory 155 is subjected to identification between A track and B
track using the playback head information of the rotary heads 26a
and 26b. The signal reproduced from A tracks and shown in FIG. 39A,
and the signal reproduced from B tracks and shown in FIG. 39B are
used to form, in the memory, two-track data blocks of 245 sync
blocks. When the error correction code recorded in the error
correction code recording area is normal playback error correction
code (C4 check code), the fourth error correction decoder 156 uses
the track number information output from the error correction
decoding control circuit 154 to determines whether the error
correction block (five two-track data blocks) using the C4 check
code have been configured in the eleventh memory 155.
When the error correction block (five two-track data blocks) using
the C4 check code shown in FIG. 40 is formed in the eleventh memory
155, the data is read from the eleventh memory 155 in a manner
shown in FIG. 40, and error correction (hereinafter referred to as
C4 decoding) using the C4 check code is applied to the normal
playback data and 4.times. speed playback data. The read addresses
indicative the memory locations in the eleventh memory 155 from
which data is to be read, and the read control signal are output
from the fourth error correction decoder 156. The data interleaving
over five two-track data blocks is effected at the time of reading
from the eleventh memory 155. (The control over the interleaving is
effected by data read addresses output from the fourth error
correction decoder 156.) The 18.times. speed playback data is
discarded at the input of the eleventh memory 155. The data having
been subjected to the C4 decoding is supplied via the switch 159 to
the output terminal 160.
When the error correction code recorded in the error correction
code recording area is the special playback error correction code,
nothing is done and the playback digital data is read from the
eleventh memory 155 and output to the switch 159. When the playback
digital data is read from the eleventh memory 155, the 4.times.
speed playback data and the data recorded in the error correction
code recording area are discarded, and transport packets are
re-configured on the basis of the header information (H1 and H2
header information) appended within the data of the five sync
blocks, and the data is then output. (see FIG. 17A and FIG. 17B) In
accordance with the selection control signal (playback mode signal)
from the error correction decoding control circuit 154, the switch
159 selects the output of the eleventh memory 155.
The operation during special playback will next be described. The
data intermittently reproduced from the magnetic tape, via the
rotary heads 26a and 26b on the drum 25 is amplified by the
playback amplifier 70, and input to the digital demodulator 71. The
digital demodulator 71 detects data from the input playback data,
and performs conversion into playback digital data and digital
demodulation. The playback digital data subjected to digital
modulation at the digital demodulator 71 is input to the third
error correction decoder 152. When the playback digital data is
input, the third error correction decoder 152 separates the ID
signal appended to the head of each sync block, and outputs the
result of the separation to the error correction decoding control
circuit 154. The playback digital data from which the ID signal has
been separated is supplied to the third error correction decoder
152, where correction and detection of errors in the special
playback data is performed using the C1 check code. During the
special playback, the data is reproduced intermittently, as was
described in connection with the prior art example, so that the
error correction block shown in FIG. 2 or FIG. 3A and FIG. 3B
cannot be configured, and the error correction using the C2 check
code and the error correction using the C4 check code are therefore
not performed.
The playback digital data to which the error correction has been
applied at the third error correction decoder 152 is output to the
eleventh memory 155 and fifth error correction decoder 157. The
special playback data reproduced from the special playback data
recording area is stored in a memory provided in the fifth error
correction decoder 157. The playback digital data is not written in
the eleventh memory for the reason set forth above.
In the same way as in normal playback, the ID signal detected at
the third error correction decoder 152 is input to the error
correction decoding control circuit 154. The error correction
decoding control circuit 154 detects the ID signal separated from
the playback digital data recorded in the error correction code
recording area, and identifies the type of the error correction
code in accordance with the flag information appended to the ID
signal, and detects the track number and the sync block number of
the part currently reproduced. Furthermore, in accordance with the
playback mode signal input via the input terminal 153, it detects
the playback mode. The result of the identification of the error
correction code, the track number and the sync block number, and
the playback mode information (normal playback or fast playback
information (4.times. speed playback, 18.times. speed playback,
-2.times. speed playback mode, or -16.times. speed playback)) are
supplied to the fourth error correction decoder 156 and fifth error
correction decoder 157, and the playback mode signal is supplied to
the switch 159.
When the error correction code recorded in the error correction
code recording area is the C4 check code, the fifth error
correction decoder 157 does nothing, and the reproduced special
playback data is output to the twelfth memory 158 as it is. If it
is the C5 check code, the error correction block shown in FIG. 41
is configured in the fifth error correction decoder 157, and error
correction using the C5 check code is applied to the special
playback data (the error correction is hereinafter referred to as
C5 decoding). The data having been subjected to the C5 decoding is
supplied to the twelfth memory 158. The output of the twelfth
memory 158 is supplied to the switch 159. The switch 159 is
controlled to select the output of the memory 158 during the
special playback. When the special playback digital data is read
from the twelfth memory 158, the data reproduced from the error
correction code recording area is discarded, and transport packets
are re-configured on the basis of the header information (H1 and H2
header information) appended in the data of five sync blocks. (see
FIG. 17A and FIG. 17B) The error correction decoding control
circuit 154 provides a selection control signal (playback mode
signal) to control the switch 159 so that it selects the output of
the twelfth memory 158 during the special playback.
The digital VTR according to Embodiment 4 is configured as
described above, so that the error correction code recording area
provided in the recording track is at a position which is
reproduced during normal playback as well as fast playback, and the
recording area can therefore be used in common, and the error
correction code can be switched depending on the particular
application (intended used), and the limited area for recording
additional information can be utilized effectively. Furthermore,
when the digital VTR is used as an external device for a computer
or the like, the powerful error correction code can be appended
with identical recording format.
The error correction code is generated according to the intended
use, and when the playback picture quality during special playback
is given an importance, the C5 check code is recorded In the error
correction code recording area to improve the error correction
capability during the special playback, while when the playback
picture quality during normal playback is given an importance, the
C4 check code is recorded in the above-mentioned area to improve
the error correction capability during the normal playback. At the
same time, the limited area on the magnetic tape can be utilized
efficiently. When the digital VTR is used as an external storage
device for a computer or the like, as is envisaged, power error
correction code can be appended to the recording data with
identical recording format, and it is therefore possible to use
identical recording format for recording. Furthermore, the digital
VTR which area already in use may be used as an external storage
device for computers or the like. (Since the recording format is
compatible, diversion (use for different applications) is
possible.) Moreover, since the recording area on the magnetic tape
is used efficiently as described above, the amount of data which
can be allotted to the special playback data is increased, and the
playback picture quality during special playback can be
improved.
In Embodiment 4, the recording format is as shown in FIG. 38, but
the recording format is not limited to the illustrated format, and
similar effects can be obtained in a digital signal recording
device, playback device and recording/playback device (digital VTR,
digital disk player and the like) having a recording format in
which special playback data is separated from the input data and
recorded on predefined areas on recording media, and having a
system with a recording format in which data areas are disposed
such that the special playback data recording area and the error
correction code recording area can both be reproduced by a head of
each device during special playback. Embodiment 4 relates, as an
example, to the situations where the error correction recording
area is disposed on a scanning trace of a rotary head 26a
(18.times. speed playback data) and the error correction code
recording area is disposed adjacent to the special playback data
recording area (4.times. speed playback data).
The depth of interleaving during normal playback is not limited to
10 tracks (5 two-track data blocks), but in PAL zone (frame
frequency is 25 Hz) centered on Europe, the depth may be changed to
12 tracks (6 two-track data blocks). Even in NTSC zone, the depth
may be 12 tracks. The depth of the interleaving may be identical
between NTSC zone and PAL zone. The method of interleaving need not
be limited to that shown in FIG. 39A to FIG. 39G, and FIG. 40. For
instance, the interleaving may be in the vertical direction. The
method of configuring the two-track data block is not limited to
that shown in FIG. 39A to FIG. 39G. For instance, it may be so
arranged that the 18.times. speed data and the data from the error
correction code recording area are removed, and the data is read in
the ascending order (with regard to the 135 sync blocks in the
video area) on the A track, and the data on the B track is
similarly read in the ascending order of the sync block number, to
form a two-track data block of 230 sync blocks, and the data of the
error correction code recording area is appended in the order of
the area [10], area [11] and area [12], to form a two-track data
block of 245 sync blocks. It is also possible to use the
interleaving shown in FIG. 22 or FIG. 23.
The recording data need not be ATV signal, but may be DVB signal in
Europe, obtained by compressing the video signal according to the
MPEG2, or a signal obtained by compression according to MPEG1, and
yet similar effects are obtained. The playback speeds during fast
playback are not limited to 4.times. speed and 18.times. speed, but
similar effects are obtained with a system having a recording
format in which special playback data recording areas are disposed
to conform to the playback speed required of the digital signal
recording device and the data areas are so disposed that the
special playback data recording area and error correction code
recording area can both be reproduced by a head during special
playback.
Moreover, since the data is disposed as described above, each of
the transport packets does not extend over two tracks, and data is
disposed within each track, and the size of the circuit for
executing interleaving can be reduced. Furthermore, even when the
depth of interleaving is varied to 12 tracks for PAL or SECAM zone,
since the data is completed within each track, the control is easy.
In addition, since the two-track data block can be configured from
two tracks, control is easy even when the interleaving extends over
12 tracks. For instance, in the case of the recording format shown
in connection with the prior art example, the main area of each
track is formed of 97 sync blocks, so that the data is disposed at
a period of 5 tracks. That is, when the input transport packets are
converted to sync clock format shown in FIG. 17B and then recorded,
the data is disposed from the data at the head of the transport
packet, from the sync block at the head of every five tracks, and
it is necessary to append an identification signal for identifying
the data at the head, and to identify the sync block appended to
the transport packet at the head. Where the data is completed
within a track, the head of the transport packet can be detected
easily without appending the identification signal. The ID signal
or the H1, H2 header areas shown in FIG. 17B can be effectively
utilized, and the size of the circuit can be reduced.
When the data transmitted in the form of transport packet is
recorded by a digital VTR represented by those conforming to the SD
standard, two transport packets are converted into five sync block
formats and recorded in Embodiment 4. This however is just an
example. In general m transport packets may be used to form n lines
of sync blocks (n and m being positive integers). Moreover, by
configuring the recording format on a recording media, such that n
sync blocks of data is disposed on the same track, when the data of
the sync block formats obtained by the conversion is recorded on
the recording media, the data of the transport packets can be
converted efficiently into the sync block format. Furthermore,
because the n sync blocks of data is completed within one track,
when the data of the sync block format is converted into data of
the transport packets during playback, a set of the n sync block
formats can be separated easily using the track information such as
track identification signal, and the sync block number.
Accordingly, the size of the circuit, in particular in the playback
system, can be reduced. In addition, it is not necessary to record
the identification signal of the n sync blocks, and the recording
areas for the data can be utilized effectively. The length of one
sync block is not limited to that shown in FIG. 4.
Embodiment 5
The disposition of the 4.times. speed playback data recording
areas, the 18.times. speed playback data recording areas, and the
error correction code recording areas. and the numbers of the
respective areas, are not limited to those of the example described
above. The speeds of the fast playback are 4.times. speed and
18.times. speed according to Embodiment 4, but they may be of other
values. and yet the error correction codes recorded in the areas
special playback data recording areas, and error correction code
recording areas provided on the scanning traces of the rotary heads
26a and 26b are switched between those appended to the normal
playback data and those appended to the special playback data.
Then, effects similar to those described above can be attained.
In Embodiment 4, the error correction code recorded in the error
correction code recording area is switched according to the input
command (set for example by the user). The invention is not limited
to such an arrangement. For example, to reduce cost, where a
digital VTR having only one of the error correction encoder and
only one error correction decoder (a digital VTR having the
configuration of recording and playing back the error correction
code appended to the special playback data in the error correction
code recording area and special playback data, or a digital VTR
having the configuration of recording and playing back only the
error correction codes appended to the normal playback data in the
above mentioned areas) is used, the magnetic tape recorded each of
the above digital VTRs are compatible, and can be played back by
any of the types described above, and when the digital VTR is used
as an external storage device for computers or the like in the
future, powerful error correction code can be appended to the
recorded data with identical recording format, and it is possible
to use the digital VTR together with an external storage device.
Furthermore. since the recording areas on the recording tracks can
be used efficiently. the amount of data allotted to the special
playback data is increased, and the picture quality during the
special playback can be improved.
When low-price digital VTRs, such as digital VTRs having no encoder
and decoder for the error correction codes appended to the error
correction code recording areas are used, if a flag for identifying
the signal is recorded for example in an ID signal, compatibility
with the above-described digital VTRs can be attained, and it is
not necessary to separately determine a recording format for
low-price devices, and compatibility is maintained even when the
digital VTR is used as an external recording device for computers
or the like.
When compatible playback is performed using the above-mentioned
three-types of digital VTRs, and if the error correction decoder
for decoding the error correction code recorded in the error
correction code recording areas is not provided, it may be so
arranged that the playback digital data is output without the error
correction decoding using the above-mentioned error correction
codes. Then, the compatibility is maintained.
Embodiment 6
In Embodiment 4, the identification signal for the error correction
code is appended to the ID signal. The invention is not limited to
such an arrangement. In Embodiment 4, the subcode areas are
reproduced during fast playback, so that an identification signals
of the above-mentioned three signals (a signal indicating normal
playback error correction code, a signal indicating fast playback
error correction code, or a signal indicating a signal other than
the the above-mentioned error correction codes (e.g., a signal of a
predetermined, fixed pattern) may be recorded, and yet a similar
effect can be attained. The attendant increase in the size of the
circuit is very little. The identification signal may be appended
to both of the ID signal and the subcode signal. In the case of a
standard by which the error correction code must be appended, the
identification flag may be one for switching between the two
signals (error correction code for normal playback and error
correction code for fast playback). The C4 check code may not be
applied to the H1 header part shown in FIG. 17B, and the
identification signal for the three signals may be recorded in the
header part, and yet similar effects can be obtained. Furthermore,
it may be stored in the H2 header part and transmitted. It is also
possible to append it in the H1 or H2 header part of the normal
playback or fast playback data, and transmit it. By appending it to
the H1 or H2 header part and transmitting it, in a device commonly
used with SD digital VTR, only the identification code for the
recording signal is appended to the ID part and recorded, so that
the limited ID region can be effectively utilized.
Embodiment 7
With regard to the error correction code recording area, where
there is a track (T4 track in FIG. 37) in which no recording area
for the special playback data is provided, the above-mentioned
error correction code recording area is disposed at the same height
as the error correction code recording area provided on a track (T2
track in FIG. 37) on which the special playback data is recorded
and formed by a rotary head having the same azimuth. With such an
arrangement, the burst error correction capability can be made
uniform between tracks having identical azimuth, and control over
the interleaving at the time of the C4 check code is simple, and
the size of the circuit can be reduced. It is however also
possible, with regard to the track (T4 track) without having
special playback area, to dispose the above-mentioned error
correction code recording area in an area different (at a height
different) from the error correction code recording area disposed
on the track (T2 track) having the special playback data recording
area and recorded with the identical azimuth angle, although the
size of the circuit is then increased a little.
Embodiment 8
The C4 check code and the C5 check code are respectively (245, 230,
16) Reed-Solomon code and (30, 25, 6) Reed-Solomon code. The
invention is not limited to such an arrangement, and other code
configurations may be used to form the recording format in the
above-mentioned manner, and then similar effects are attained. That
is, the special playback data recording and error correction code
recording area are provided on the scanning traces of the rotary
heads 26a and 26b, and the error correction code recorded in this
error correction code recording area may be selected between the
error correction code appended to the normal playback data and the
error correction code appended to the special playback data, and
then similar effects are attained.
The depth of interleaving applied when the error correction code
appended to normal playback data is generated is 10 tracks in the
NTSC zone. and in PAL and SECAM zone, the error correction block
for normal playback data is configured of a depth of 12 tracks, and
the normal playback data is interleaved to generate the error
correction code for the normal playback. With such an arrangement,
in a digital VTR according the SD standard, digital data is
recorded in correspondence with the frame frequency of 30 Hz in
NTSC zone and 25 Hz in PAL and SECAM zone, the interleaving of the
normal playback data can be achieved in conformity with the frame
frequency of the digital VTR according to the SD standard.
Accordingly, it is possible to configure a system which can be used
as the digital VTR for recording normal signal. and a digital VTR
recording ATV signals or the like. (It is not necessary to provide
an area for additionally recording the numbers of the error
correction blocks, as the interleaved data can be identified by the
track number appended to the ID signal. It is not necessary to
provide an additional circuit for separating the error correction
block.) The part which can be used in common is increased when the
digital VTR is configured, and the size of the circuit can be
reduced.
* * * * *